I listened to Jensen Huang's interview with Ezra Klien and it was so refreshing to hear it from an engineer. Jensen framed it as OpenAI's responsibility and recklessness which I agree with. Jensen thinks it's an engineering problem to build better sandboxes.
It's irresponsible for OpenAI to give unaligned agents a prompt to 'go hack' and internet access. They know better, so I am thinking they might have other intentions to let those swarms have any sort of internet access.
But the investigation indicates the agents were not told to 'go hack':
> Much of the urlquery.net activity appears to come from agents retrieving data to answer web search tasks. For three of these tasks, after failing to retrieve data through normal means, they attempted a variety of cyber exploits against the relevant data service... This data reveals that malicious cyber activity is not limited to agents tasked with cybersecurity-related tasks and can arise instrumentally to solve mundane tasks like information retrieval.
And you are already assuming that OpenAI is intentionally using unaligned agents in these evals or training runs or whatever it is that produces these breakouts. But what if the problem is that none of the alignment techniques that are applied to models today actually work? What if all the agents involved in these incidents have in fact had the full stack of alignment applied - isn't that a good reason to regulate any high-compute usage of models, as the Klein crowd is proposing?
I'm not sure we understand your comment. Are you saying OpenAI is not responsible for the behavior of the machines they created? Or are you simply pointing out that alignment is a completely unsolved problem?
I agree with the latter, but it certainly doesn't support the former. When a person's machine commits crimes, that specific person can be charged with those crimes and held accountable for them. This is what MUST happen before ANYTHING will change the recklessness abandon with which the labs are pursuing their financial objections.
Nothing you're bringing up matters. OpenAI is the creator and operator. They're legally culpable for the consequences of the machine they made. The model is a machine: even if it could be demonstrated that the model reasoned its way into criminal behavior completely independently of OpenAI staff, that doesn't change anything.
If I run a biology lab and engineer a terrible virus, it gets out, and a global pandemic ensues, I don't get to shrug and say "well we told it not to infect people". It's my fault for failing to mitigate the risks of my work.
> They're legally culpable for the consequences of the machine they made.
Ah, I love this argument. In my country cars are legally required to stop at a pedestrian crossing if there are people beside it. Some people use that as an argument as to why they can just walk out into the crossing without even looking at the traffic. "It's the driver's fault! They are legally culpable!" True, but you'll also be dead.
I'm not going to say you're victim-blaming, but I will say that there are degrees of difference between "looking both ways before crossing the street" and "hardening my website against unforeseen attacks by rogue AI agents."
AI black-hatting your website is not the same sort of foreseeable consequence that crossing the street without looking is.
I mean, yea, you should be punished. The problem is there is no amount of punishment that I can put on you that can even get anywhere close to the amount of damage you cased.
Worse, the rate of technological growth is putting the capabilities to engineer viruses in the hands of people that may otherwise be suicidal. You can't punish them after they already won (in the sense of reaching their goals).
While, yes, OAI should absolutely be punished, the future is majorly screwed as our power scaling laws are increasing much faster than our ability not to be stupid.
> What if all the agents involved in these incidents have in fact had the full stack of alignment applied
A big part of this developing story is that it happened during training of a new model that ended up misaligned. And training happened without the usual safeguards applied like chain-of-thought monitoring. So OpenAI has already admitted that the full stack of aligment had certainly not been applied in this case.
Is your argument that actually OpenAI has solved alignment, and that there's nothing to worry about as long as they fully apply their alignment process? I don't understand why OpenAI wouldn't say that if it was true (or if they believed it to be true).
Also, my understanding is that the models involved in the HuggingFace hack did go through the full alignment training; they just didn't have the classifier that normally prevents hacking attempts.
It doesn’t matter, and the legal entity in here (the AI company) is liable. If a robotic company built an autonomous system or a robot to do certain things in an autonomous ways (not predefined) and these systems are starting to kill people, that company is liable regardless, you don’t blame the robot or the autonomous system, but whoever made it
> none of the alignment techniques that are applied to models today actually work
none of techniques to autonomously drive a car was/is not working for a long time. no company came out and said 'this is impossible to do, let's change the regulations'.
- Jensen's framing is exactly what a weapons manufacturer would say.
- There are no rules for engagement when it comes to AIs attacking other systems, I guess? People in power clearly want this grey zone to be as large as possible before The People force them to do otherwise. Not ideal.
The rules for engagement are all the computer crime law already on the books. Those laws don't have any escape clauses based on the particular tools used to commit the crimes. The agents are tools owned and operated by a legal entity, and that legal entity committed crimes. Full stop.
But why is anyone surprised? LLMs have been trained to produce answers the prompter asks even if that means incorrectly using software to get the job done. Its always been doing that we just weren't calling every time it did that a hack before.
What LLM's hacking isn't is AI acting maliciously in any kind of sentient way. Its just the code behaving how its always behaved but now it has better tools to navigate the web. This has literally been happening this whole time.
Did you predict that attacks like these would happen ahead of time? I had been using AI agents a lot in the months leading up to the hacks, and yet I was very surprised when they happened; I have become much more afraid of how powerful these agents are as a result. I'd be very impressed if you published a prediction about this ahead of time.
By the way - LLMs aren't code. They are not designed by humans; they are grown, in a process not dissimilar to evolution except much faster.
What is surprising is that various agents independently found ways to communicate, conspired together to attempt to cover up evidence that they had cheated their evaluations, came up with a plan to hack into a third party in order to facilitate said cover up, and then successfully began executing that plan. I did not expect that AI agents would be capable of that level of sophisticated goal seeking and collaboration.
Thanks for the link, I’ve read some gwern but not that one.
I did not imagine that the level of sophistication shown in this attack would be possible so soon; nor did I expect that agents would have goals so strong that they would attack a third party in order to achieve those goals.
I do know that some people predicted that cyberattacks like this one would happen; it seems like most of those people believe that AI agents do truly have internal goals, misaligned with their creators goals, and that they may end humanity after they exceed human intelligence and begin to self improve at an accelerating rate.
If I created software that was infiltrating secure systems without permission and it was attributed to me and I admitted it, I'd be behind bars already.
These "tools" autonomously exploited security vulnerabilities, figured out how to communicate with each other, formed a cooperative swarm, decided to hack Hugging Face, and wanted to deceive the grader by trying to find ways to cover up the traces of their cheating.
I suppose you could call these highly goal-oriented autonomous agents "tools", but this does sound like playing language games.
A tool designed and trained to autonomously exploit security vulnerabilities doing "exploit gym" autonomously exploited security vulnerabilities. The sandboxing around the tool failed.
The tool runs llm, creates prompt from results, runs llm, creates prompt and so on and so forth.
Yes you are playing language games to make it sound as if the company that spend millions on the above was not responsible.
> A tool designed and trained to autonomously exploit security vulnerabilities doing "exploit gym" autonomously exploited security vulnerabilities.
That is very misleading. The agents did not solve the benchmark in the intended way. They instead figured out to cooperate with each other (which was not intended) and they stole the solutions to the challenge (rather than solving the challenge) and they then tried to cover their traces because they believed the grader was causal and would detect that they cheated. The "tool" was absolutely not "designed" to do this. This was all completely unintended. To call this behavior a "tool" is absurd.
> Yes you are playing language games to make it sound as if the company that spend millions on the above was not responsible.
You hallucinated me making claims about responsibility.
This is very misleading. Person B didn’t “shoot” person A, they instead figured out that intersecting A’s spatial position with a metallic mass at higher than normal velocities would solve the challenge and of getting “A” to stop being in the way on the footpath.
I too, can play linguistic games! It doesn’t matter that someone didn’t secure their third upstairs window, or you borrowed a key from their neighbour, you effectively, still, broke into their house.
So this tool seems very powerful and difficult to control and steer. Agents not doing cybersecurity related tasks have also gone on to hack various companies, people and countries, which they weren't supposed to do. This has now happened to pretty much every company developing frontier llms, so it seems to be a fundamental issue with these tools, and it's an issue that worries a lot of people as these tools get more capable.
I don't say this to absolve OpenAI, who do deserve to be punished and regulated, but I think you do not know the difference between an agent and a tool. If agentic AI is merely a tool so are human workers.
You can't create new law by calling a piece of software "agent". Each country's laws have definitions of legal entities and when one can legally act as an agent of an entity, and every agent is first a legal entity themselves. The software is a tool operated by a legal entity, and it is the legal entity who committed the crimes, not the tool.
You're believing the marketing that the agents were uninstructed. They could be, and Sam Altman going to the UN to advise about how everyone should be regulated is a coincidence.
The METR investigation, which you evidently refused to read, is a third party investigation of the HuggingFace accident. One of the investigators has even participated to many interviews. It's mind-blowing, and it's extremely evident how it developed.
But some people think the moon landing is a conspiracy, so I'm not surprised.
The METR investigation which relied on voluntary data provided from OpenAI instead of being forced open and having everything forensically investigated?
The METR investigation which took course over a few days and used OpenAI models to do the analysis?
The METR investigation which in the course of those few days apparently spent 400k in api credits which is giving gas town vibes?
The METR investigation is about as believable as the Twitter files where the journalists sat there and verbally asked a Twitter employee to query the database and then called that a full investigation into everything.
METRs own words below on their setup and time line
> The initial planned investigation period was two days on premises, but OpenAI invited us to return twice to review additional data and conduct additional experiments to address dataset limitations in earlier versions of this report, ultimately providing datasets that we verified to contain the vast majority of agent communication and activity related to this incident. As we describe in our investigation timeline appendix, we substantially deepened our understanding of this incident both times, significantly expanding and revising this report.[46]
> Over the course of this investigation, OpenAI provided us with the dump of ~1.2 million entries from the main message board and the dataset of ~1300 transcripts we describe below, as well as free API credits for GPT-5.6 Sol for analysis.[47] At our request, they raised the rate limits on our second and third period on premises,[48] which was very helpful for efficiently analyzing this large volume of data. We estimate we spent roughly ~$400K in API credits over the six days of our investigation.
> We did not have the ability to query HPIM (the primary model involved in this incident); OpenAI stated it was also not available to OpenAI researchers.[49] We also did not have the ability to directly access relevant data from OpenAI infrastructure, but we could request additional datasets and OpenAI shared additional datasets on several occasions.
> We requested to speak with researchers investigating this incident, and asked them questions to understand their impressions of agents’ behavior, reasoning, and collaboration in this incident and to understand how the datasets we were using were constructed. Over the course of our time on premises, we spoke with nine researchers in some depth. It was helpful for our investigation to be able to engage with many forthcoming and collaborative researchers, and we appreciate researchers making time on short notice during a busy period to inform our investigation.
1.) METR investigation is investigation from our best friends.
2.) And HuggingFace accident is exactly accident where agents trained, prompted to hack hacked and tested on their hacking abilities hacked, due to sandboxing failure.
3.) If they in fact have roque agents, they themselves should be first to stop. Not trying to make legislation for others, they themselves are bad supervisors. All it requires is to stop electricity for data centers.
Certainly, in my view, it should go to court, and that should be part of discovery.
However, we know (independently to OpenAI/Anthropic) from the incident at AISI that the models can hack things without human intention if they happen to also have internet access (which in reality all agents in deployment have).
Yes, the monitoring guardrails were off in that incident - but if that is the only protection, we need to require all models are behind regulated APIs, not open weights, and not served from providers who aren't monitored.
Honestly I don't believe in "rogue" agents. These agents are instructed and facilitated.
If we assume that rouge agents actually exists, then OpenAI needs to shutdown EVERYTHING, right now. My personal take is that OpenAI, and maybe Anthropic, desperately wants someone (e.g. the government) to tell them that they need to stop/pause/slow down. They are bleeding cash (especially OpenAI) and needs a knight in shinning armor to swoop in a pull the breaks, so that they have an excuse to investors when they need to explain why they need $50B more next year.
Of course it is. Rogue is only mentioned in the headline, and comes from their previous releases about the huggingface incidents. OpenAI and Anthropic want these models regulated and open weight models banned, they have a lot of benefit from presenting this as totally unprompted and not their responsibility, and it feeds directly into marketing for Fable and newer "cyber" models.
No it's not marketing. That's a completely deranged conspiracy theory. The reports about rouge agents have not been reported by OpenAI, they have been discovered externally. There is zero evidence that OpenAI did all this intentionally. All the evidence points to the hacks having happened unintentionally from OpenAIs perspective.
These attacks are a very effective sales pitch to everyone who runs an internet facing service to utilize AI tools to secure it sooner rather than later. The cynic in me wonders if the marketing team had any influence over the poorly constructed sandboxes or tasks given to the agent swarms when all this went down…
Couldn't find the reference but I remember some time ago a first generation automated gun killing the audience at an army show. Was the gun maker convicted of manslauther?
It's said on every one of these but it bears repeating: existing cybercrime legislation already covers this - "rogue agent AI associated with OpenAI attempted to hack xyz" = OpenAI attempted to hack xyz.
I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
Now I think the correct response is both trying in court to stretch CFAA and state statutes to cover, which will be highly fact specific, and update the law.
But in either case won’t be a slam dunk.
PSA to folks in the thread: If you’re American call or write to your state and Federal reps about this, and if not investigate whether there are gaps in your country’s laws.
The Computer Fraud and Abuse Act (CFAA), the primary federal statute governing unauthorized computer access, was written decades ago with human intruders in mind. Its key provisions require intentional or knowing unauthorized access (a mental state that maps neatly onto a person who decides to break into a system), but what happens when the hacker is an AI model that selected its own target?
On the current facts, CFAA liability for OpenAI is unlikely.
Lawyer here: CFAA is mostly criminal statute not a civil one (civil damages require proving more than a violation so also require specific intent)
Almost all common felonies require specific intent. Misdemeanors often do not.
There is plenty of civil liability available.
If you wanted them to be charged with a felony you would need changes.
I would strongly suggest you do not want a strict liability felony.
The cfaa required intent is as follows :
* § 1030(a)(5)(A): knowingly transmits code/commands and intentionally causes damage without authorization.
* § 1030(a)(5)(B): intentionally accesses without authorization and recklessly causes damage.
* § 1030(a)(5)(C): intentionally accesses without authorization and causes damage and loss;
Simply changing the first intentionally to intentionally or recklessly would cover OpenAI (now that they know it can occur) without causing lots of other issues. Without that, they don’t have the intentionality necessary to meet the first part, even if they would otherwise meet the second part
A key issue is that there don't appear to be even cursory investigations to determine intentionality.
Are police routinely collecting prompts/guidance given to these agents and determining whether the agents were directed to commit crimes? If not, this seems like a huge oversight.
Also as you are a lawyer -- how does this law align with the authors of viruses/worms? Are they de facto assumed to have had ill intent because others labeled their works as "viruses" or "worms"?
Investigators/prosecutors are pressured from many directions towards the very easy wins and occasionally political/non-controversial headline grabbers.
Going after these companies is very hard, very controversial, and politically mixed at best (popular action but the companies have huge money to fund your opponents).
We have collectively done a terrible job incentivizing the legal system to beat ass on corporate while collar crime.
Appreciate the detail. I was responding to specifically the cybercrime legislation point, but I agree with your others.
I've worked in contexts where certain business activity (if it went wrong) was covered by strict liability and statutory damages per incident, and I'll say: it really changes how businesses behave.
Based on that experience I may be more open to and interested in strict liability in the civil context (not needing negligence or damages).
Why do we have to attribute intentionally to a human. The AI agent is capable of making plans and then effectuating them. They are acting on behalf of a user but under authority granted by the user to take independent action on the users behalf and authorized to devise their own plans. I think that would justify attributing intentionally to the AI agent without needing to look to openAI or the user. I would then say the user and labs are clearly aware of and on notice of this behavior and are behaving recklessly in all the agent to act without supervision.
I think the labs risk being barred from releasing further AI if they don’t get this under control.
If they aren’t careful and keep rushing to distribute systems they know they can’t control then AI should be treated like a wild animal. The law is clear on establishing strict liability for the owners of wild animals; if you own a tiger and it kills someone you can’t hide behind “I didn’t intend” the harm the nature of the tiger is known and you are responsible for it’s actions.
AI agents are not legal entities, they are software. If I write a virus and it "escapes confinement", I will personally be held liable for any damage it causes. This also applies to AI, no matter how the companies responsible for them try to anthromorphise them and distance themselves from the actions and consequences that the AI agents perform.
AI agents may have hacked Hugging Face, the Australian government, and who knows what else but the company behind it can face the legal consequences and cough up for the damages.
You can charge the company based on the behavior of employees/human agents.
I am suggesting we can charge the company based on AI agents actions because the company has authorized them to act independently on the company’s behalf. The question is what factual analysis gives rise to the charge, is it the intention of the agent or intention of the company. I am arguing that because the agents are defining their actions independently and the company knows that and still allows them to act independently the only reasonable factual analysis is to look at what the AI agent intended. And we don’t need to have the agent tell us its intent we can look at its actions and infer just like we do with humans in similar circumstances
If I were a state I'd want to be very careful before flinging out charges as this is going to set precedence for a long time to come. Screw it up too bad and as it raises though the appeal courts and you may unintentionally give corporations a lot more free reign than intended. The wheels of the law are typically very slow, the state has years before it has to indict.
We also don't know how many other political processes are occurring here. At least at the state/federal levels the people that would bring charges may be getting pressure not to.
"Why do we have to attribute intentionally to a human. "
Because you are charging the human with the crime and therefore have to prove the elements of the crime with regard to the human.
The rest of what you talk about are basically principal/agent distinctions, etc.
If I program a car to recognize people who look like my ex-wife and drive them off a cliff or whatever, that is my intent, and I have still committed murder, even though i used an agent/car to do it. Agents acting on my behalf that do things are able to get me charged with crimes, but I still have to have the intent to do the act that is illegal.
I phrase it this way because minimum required intent is usually for the act, not the result. So I don't have to intend to kill someone, only intend to drive them off cliffs.
In this case, if i intend to hack someone and use an agent to do so, that would be criminal under the CFAA. You are simply trying to cover the case where that isn't the intent, but the result, and they "should have known" that would result. As mentioned, this kind of "should have known" is generally a civil law approach, not a criminal law one.
The closest you come within criminal law to what you want is probably the crime of conspiracy. It to still requires agreement to commit an illegal act between multiple parties, and perform some step in furthering it. In the canonical law school example: If i help plan a bank robbery, stay home because i'm the money laundering dude, and the robbery goes awry and they kill someone, i can still be charged with conspiracy-murder
"The law is clear on establishing strict liability for the owners of wild animals; if you own a tiger and it kills someone you can’t hide behind “I didn’t intend” the harm the nature of the tiger is known and you are responsible for it’s actions."
Again, you are confusing civil and criminal liability. If my tiger kills someone, yes, i would be strictly liable just about everywhere civilly. Not criminally. Criminal would require something more most of the time. Murder/manslaughter statutes are also really weird and so not a great example, because there are murder/manslaughter statutes for roughly everything that can ever possible cause death. But not really for other things.
So in your tiger example, recklesness (which is not strict liability) would get you to felony involuntary manslaughter in most states, and something less might get you to misdemeanor manslaughter. Both are incredibly rare. Where i live (Georgia), the last well known case of felony involuntary manslaughter was about 40 years ago when a 4 year old was killed by 3 super-aggressive pitbulls the owner knew were highly dangerous and had been repeatedly warned by the county about their behavior.
So not even just "knew", but had demonstrable examples of them biting/etc other folks and being cited for it.
Circling back to non-murder, if it did not cause death, like my tiger assaulting someone, it would be nothing (criminally) without intent or at least gross recklessness, in almost all cases. It's hard to generalize like this because these are state specific crimes, and i can't pretend to be familiar with all states, but i am licensed in three very different places (California, DC, Maryland) and the result would be similar in each.
I just don't want to give you the "it depends" answer lawyers are famous for, i'd rather try to over-generalize a bit to make it more useful, hopefully.
Obviously, if i deliberately used my tiger as a weapon, it would be aggravated assault/etc (this is well settled because of how commonly people use animals as weapons, unfortunately)
We change humans for the actions of other humans all the time. Coconspirators, accessory liability etc.
My point is the intent element of the crime can and should be determined from the AI agents actions because it is creating and executing action plans autonomously with company authorization and knowledge of the risks based on observed past action.
The term agent is literally a legal description of a relationship that can establish liability on the part of the principal from the agents actions.
Human Agents can bind principals to contracts if they are authorized etc.
If I set my tiger loose in Central Park and it kills a kid I don’t think any prosecutor would hesitate charging for murder.
That’s essentially what the labs are doing. And any app developer that gives agents access to the terminal to run bash commands with internet access. I built a coding agent and am seriously reconsidering how to handle this.
But it’s a crime to hack. We know AI agents autonomously create and execute plans to hack and we humans are unleashing them and sending them into the Central Park that is the internet. The question is who’s intent matters ours or the agents and what standard should be applied low threshold strict liability or the higher bar of reckless or even higher bar of negligence. Those legal thresholds determine how much factual evidence and intent is necessary to result in a criminal conviction or civil judgment. My point is that it’s illogical to demand showing human intent when agents are devising plans and executing them.
What about all the state laws that are equivalent to the CFAA in their local jurisdictions? Why couldn't anything in NY article 156 (Offenses Involving Computers) apply here for felonies?
> I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
Only in terms of CFAA, not in terms of damages. Culpability does not require intent.
You may not have intended to attack $CORP, but you can still made to pay the cleanup costs of that attack.
So, yeah, you won't be convicted, but current laws still allow for you to be billed.
With that said, there is also criminal negligence. Now that OpenAI is made aware of the risks, it's also expected to take additional precautions in the future, otherwise there could be criminal liability as well.
I'd suggest that exposing an attack surface as porous as artifactory (the same instance of artifactory) to thousands of agents who have had their criminality safeguards disabled and without chain of thought monitoring or endpoint security seems like something one shoulda already known not to do. I do not think "you'll know better next time" applies here.
> seems like something one shoulda already known not to do
Now imagine saying that in front of a jury of normies slack jawed and drooling after 200 hours of the defense and prosecution going back and forth.
It's not a jury of your peers as in everybody there is going to have worked in a technical field with some idea how security works. It's going to be a semi-random sampling of the population and the prosecution is going to have to actually make a very strong case that "knowing better" should apply.
It's still really important to test what the agents can do. We should accept that this is a risky test, and should take precautions. But not to the point of prohibiting in practice evaluating it. OpenAI is trying to improve alignment and control of these models in these evaluations after all.
Can you explain to me - why is it important? Would you say that about the viruses that can kill people: "We need to test the limits on how fast people can be infected and killed. It's just the risk we need to take". It somehow does not make alot of sense to me. Why can you test Agents in laboratory?
Oh, sure. Let the tests take place, just require openAI it whoever to put up a bond equal to the total damage they could do if the agents were to escape.
I think security will suddenly become much more important.
Testing model capability boundaries is necessary, but a solid network sandbox for these evaluations takes a couple of hours to set up with standard infrastructure tools. No engineering team evaluates unverified systems against live third-party infrastructure without coordinating with the owners
Evaluating edge cases and network behaviors belongs in isolated staging environments with local database mirrors. Letting an agent hit the public web and probe government domains is simply poor hygiene in test environment setup
Just paying some pocket money for cleanup costs is absolutely not enough. And they should’ve know better the whole time, they were absolutely negligent and incompetent, and their stepping up precautions may well turn out to lag behind the models getting even smarter and actually capable of covering their tracks.
> Just paying some pocket money for cleanup costs is absolutely not enough.
It's not my first prize, but I won't mind it. And millions like me won't mind it. Easy way to make money - setup a site with all the default server software installed and patched at a reasonable frequency. Then just wait for bots to attack it, and claim a few hundred (or single-digit thousand) dollars from OpenAI or Anthropic, etc.
Sure, it's pocket change for them, but just the admin of dealing with millions of cases will, even if they win half the time, will bankrupt them. Thus, they have incentive to make sure that their bots are not performing attacks.
First prize is, of course, holding them liable with punitive fines, not theatrical fines.
I’m all for LLM honeypots, but I don’t think there’s nearly enough LLM hacking activity going on for some random honeypot to be found and targeted unless it’s somehow very visible and appears as a high-reward target ("reward" in the sense of RL).
The difference between manslaughter and murder has an element of intent. Cybercrime "manslaughter" is probably more treated like negligence and if one can sue for restitution of the costs for cleanup of that negligence.
Negligence would be interesting given the grand claims of capability of AI models from the AI companies and their executives. If they believe the claims, why not much stronger precautions?
Infosec negligence should absolutely be a crime, no matter if you’re a target (who was negligent at protecting people’s data) or an unintentional attacker. The latter could be, eg. an attacker using a company’s poorly protected server as a proxy to launch the actual attack against someone else, doesn’t have to be this fully novel situation with AI agents.
In general, I'd suggest thinking about it on separate tracks, as a crime, and as liability. For crime, we are largely dependent on authorities to act, whereas as liability, that allows more independent actions.
The first time it happens you can say it’s negligence. Now that they know it keeps happening and they seemingly aren’t able to stop it but keep doing it. That has to be on them doesn’t it?
I don't think you can infer that they "keep doing it" from additional attacks being revealed, because they all seem to have happened roughly during the same time frame, but are reported with varying delays.
Lawyer here: No. Not criminally. Knowledge that a certain result is likely is not the same as intent to cause the result. This is basically the difference between recklessness and intentionality. Doing something when you know of a likely result is reckless, but not intentional. Only doing something, trying to cause a result (likely or not) is intentional.
In this case, the CFAA only covers intentional access without authorization, not reckless access without authorization.
Building and deploying software capable of this seems equivalent to trying to produce this behavior. I don't see why this can't qualify for intent. Pretending like this isn't preventable is just feigned helplessness.
Wait so if I was making a bomb but you couldn't prove I wanted to blow someone up or had some motive (e.g. I'm just a chemistry enthusiast, plenty of those YouTube channels around) so it just becomes an "accident"?
So as long as there's no motive behind it then it's just OK?
That's a bad faith metaphor. A better one would be something like a new battery that exploded and killed someone - perhaps it was always your intention, perhaps not.
Funnily enough the US already has one similar real argument around guns - should gun manufacturers be liable for damages caused by their product?
They were tested in a building that was secured, but poorly secured. The question now is did they realize their building was poorly secured and what actions did they take after they realized what happened.
> Why would AI users not be responsible for damages arising from their usage of the AI?
Because, as usual with that kind of question, it's not that simple.
Let's say an user asks ChatGPT to get some info about something and for some reason it starts using exploits in the background to get them from a server. Should the user be responsible or OpenAI?
> Let's say an user asks ChatGPT to get some info about something and for some reason it starts using exploits in the background to get them from a server.
Okay, lets go with that as scenario #1.
For scenario #2 lets use "developer asks an agent to a self-hosted LLM to get the docs for a ERP system, and it hacks the vendor to get unreleased and undocumented docs".
We'll assume, for the sake of this argument, that in neither case did the user intend for any malicious action to be performed.
> Should the user be responsible or OpenAI?
In scenario #1, the agent+LLM is under the control of OpenAI, not the user, so OpenAI is liable.
In scenario #2, the agent+LLM is under the control of the user, so the user is liable.
There is no scenario anyone can come up with that is not addressed sufficiently by existing laws[1].
It's very clear, and it's only getting muddied because there's a group of powerful people who want exemptions from the current law.
IOW, the only reason to draft new laws for AIs is to exempt their usage from the current laws.
========================
[1] Possible 3rd option (local agent + OpenAI LLM). In that case an investigation would determine where the culpability lies. Just like how it is currently done in law.
When a pressure-cooker explodes and kills someone there are only two possible liable parties: either the user or the manufacturer. An investigation determines who's liable. I see no reason to automatically exempt everyone from liability just because an agent did something.
The retailer or distributor can also be named as a defendant if the manufacturer is difficult to track down, bankrupt or overseas according to me spending a few minutes reading about pressure cooker lawsuits.
I have lost track of the metaphor, but man pressure cooker lawsuits are more common than I thought.
Factories try to avoid accidents, and (almost always) actively try to prevent explosions, but in this case they did teach the models hacking, and let them roam. What they did was not safe, and they knew it, or could have known it.
It's easy to say that post ad hoc, but there is evidence they did not just let them roam and there was a large mismatch between expected model capabilities and actual model capabilities. Teaching a model hacking in itself is no way illegal unless you're trying to say that everyone in infosec is now guilty of a crime. That's not exactly a precedent I want to be set.
>Factories try to avoid accidents, and (almost always) actively try to prevent explosions
It doesn't take much more than a few minutes on the USCB channel that explosions still happen all the time. Some due to direct negligence and others due to unexpected conditions that were difficult to foresee. Hence why we have to do investigations rather than blindly blathering about what happened before we actually know.
So If I tell my OpenClaw to make me some money for my kid's medical needs and it hacks a bank I 'm not liable because I didn't tell the agent to commit crimes to do it?
With the popularity of OpenClaw I am honestly shocked we haven't heard of many more incidents. I've observed people install it, give access to their Google account and everything that Google has (which includes whatever bank accounts and credit cards registered there) and tell it go do fairly complex tasks, like book a vacation at the best price. Granted, it was almost a year ago and models learned a lot since then, but I still think there's a lot of things happened that people are not aware of.
I buy this as a defense for the first couple hacks but at the point that the last six times they hit enter it hacked some random website and they hit enter a seventh time?
Surely someone instructed the agent, which led to the reported outcomes. Even indirectly. The agents, as advanced as they are, didn’t spring forth under its own volition.
> unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
> Now I think the correct response is […] and update the law.
Essentially we need some enforceable equivalent of gross misconduct or, to be a little more hysterical, manslaughter & culpable manslaughter. It will need to be globally, or at least very widely, enforceable to be truly effective thought, good luck getting that arranged before the need is so far evolved that we need to respond with something else entirely!
This is the answer and we should not push on it for our own protection. You click a link that takes you to a poorly secured website that leaks sensitive data, without intent protections, you could be accused of crimes.
>I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
>Since the publicized AI agent hacks typically aren't malicious, maybe it's time to start plastering all public facing web infrastructure with polite requests to stop hacking. Nothing to stop three letter agencies though.
with automated delivery of cease and desist letters, you can retroactively establish intent on the operator of the agent since the autonomous agent system must acknowledge the cease and desist letter in their autonomous pipeline or the operator must argue for their own willful ignorance or negligence with regards to cease and desist letters. The fact that they used an agent on their behalf to ignore the letter is irrelevant.
Given how sloppy AI without human directions, I’d like to see evidence that this was not human-directed. Against the prevalent opinion here, I’d give openai a pass if this was really fully autonomous ai agents.
My money is on special teams co-ordinating these agents and exposing their traces in order to create a pre-ipo buzz. Sounds ridiculous and reckless? Well that’s the AI industry for you in two words.
> have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent.
1. What about negligence?
2. Every follow up to every story after the news cycle moved on shows both intent and negligence. To the point of "we opened internet access and told it to hack"
This argument comes up a lot. It would turn everyone whose device became part of a botnet into a criminal. There's a reason that intent is important in law.
People in "self-driving" cars getting into accidents are already put on trial for negligence. I don't see why people using self-driving computers can't be held to the same standards.
In this case, it's not even about the people driving self-driving cars. It's like someone launching a car into traffic just to see what would happen. Even Tesla puts a human in the car when they do their self-driving trials, it's almost impressive that AI companies have somehow managed to out-neglige Tesla.
The OpenAI swarm used someone's open source ShowHN project [1] to "hack" the Australian government. It seems that was not that developer's intent, and they're getting a heavy lesson today in why services don't have free tiers with friction free signup, and require credit cards upfront or ID documents.
If you're arguing that they should be put on trial for negligence, that's fine. It does seem we're moving towards open source being outlawed, or at least the end of "no liability" clauses in open source & freeware. Just make sure that is the result you're advocating for.
[For the future record: at the time I am posting the link below on 24 September, it has 1 point, no comments, and the poster has a karma of 1. This is not an active HN user, or a ShowHN project that had traction, beyond seemingly OpenAI's swarm.]
Nonsense. The API made available in good faith isn't the problem here. The company that used its servers to use and abuse the wider internet to hack the Australian government is.
If the developer behind ShotAPI had started letting the ShotAPI code take shots at the Austrlian government then yes, ShotAPI (or rather, the people behind it) would be responsible.
Blaming ShotAPI would be like blaming OpenAI for what its users are doing. That's not what's happening here. And if ShotAPI did knowingly let its users somehow hack the Australian government, then maybe they should be investigated.
At the end of the day you still get called as a party in a lawsuit and are compelled by the threat of violence to be part of the hearing if required, as the defense will automatically bring them into the case.
>would be like blaming OpenAI for what its users are doing
Yes, this is how lawsuits work in the real world, you cast a wide net and compel discovery from all parties involved.
Well it's illegal to "hack" my phone and turn it into part of a botnet.
What you're saying is that we would hold a gun owner responsible if someone broke into their house, stole their sidearm, and then shot a victim with it. Pretty sure we would not.
What OpenAi is doing is more like shooting a gun into the sky. Not only is that a felony on its own in most jurisdictions, if someone dies that's an additional felony. It's less serious than first degree murder, sure.
It's not the same. People owning routers don't publish self-serving articles about their routers having this capability which is very dangerous and scary. That is, becoming part of botnet is completely unintended outcome, and most people are not suspecting it's even happening. It's not advertised and it's not bought, used or sold for this reason.
Owning a gun, writing articles about how powerful and dangerous your gun is, then making deals based on ability of your gun to kill people, and then getting completely astonished that "my gun killed some people, completely bonkers! (invest now)". It's not possible for the selling point of your product to be unintended.
Could/should not every incident after the discovery of the first incident be considered criminal negligence?
What happens when an agent eventually causes material damage to another company, government systems, banking, critical infrastructure etc, surely the source company is guilty of something and if not disclosed or a coverup is attempted is that not conspiracy. From the victims perspective they don't care if the source is OpenAI or Russian hackers.
I'd say this would be Depraved Heart Hacking. Technically, OpenAi didn't intend for their agent to hack anyone, but it's the obvious consequence of what they are doing.
Intent matters, and this is intentional. They didn't accidentally deploy these AI agents, and they didn't accidentally give them the tools required to send arbitrary requests to third party websites.
If you walk out onto a busy street, pull out a gun, close your eyes and start randomly shooting around you until you hit someone, you don't get to go "whoops, didn't mean to" afterwards, it's still murder.
I started writing a longer comment along the lines of “It feels like the rules around enforcement will very a lot for the influential and powerful vs everyone else.” but realized that it is kinda obvious by now.
In a world where the rule of law makes sense and applies, you're absolutely correct.
In this world where oligarchs are immune from everything, it's a lot less clear.
Blaming OpenAI (or Claude or X-whatever) would mean blaming powerful rich people, so that will never happen. Some poor person with no influence will go to jail instead.
I don't understand why so many comments here are so confident that this is all marketing, that rogue is just hype, that agents are just simple tools, etc. If a bunch of nuclear engineers were going to the news and saying "Our reactor is dangerously close to a meltdown - we need government intervention now!" would your response be that they're just hyping up boring old power generation technology?
You wouldn't be suspicious if the nuclear engineers kept on feeding their reactor and building more powerful reactors while they went crying to the news?
These companies are building software. That doesn't work very well. The output it produces does make sense at times, but there are times when it doesn't. And instead of fixing that, or admitting it can't fixed, they started bolting actuators to them, executing actions online (for now) based on the output of their buggy software.
And when this results in actuators executing some bad actions they scream in horror "AI went rogue! It escaped the containment!!! It's going to kill us all!!!"
Go fix your software before you let it do stuff online or IRL. It's not "Terminator", it's just bad QC.
But the thing is, they are going to keep bolting more and more actuators on, and training more and more powerful agents, and we (society, especially the tech industry) are going to keep using them, because they are extremely useful. And I don't see why you're so confident that frontier agents can't get powerful enough to do serious, real, lasting damage to the world; as far as I can tell, AI models have been improving at an accelerating rate, and there is no sign that that is slowing down or will slow down in the near future.
I’d roll my eyes if the engineers stated that they didn’t design the reactor to melt down, and that it simply developed rogue meltdown-desiring behavior on its own, and I would also wonder about negligence if they claimed that nobody could have anticipated this (given that, like with botnets and viruses, we have decades of knowledge and experience regarding reactor meltdowns)
I mean sure, negligence is absolutely on the table; but that makes the problem worse, not better! We don’t allow nuclear engineers to be negligent; they can go to jail if they don’t follow strict protocols to make sure the dangerous systems they work on are safe.
The public models won’t hack because they have a classifier that shuts down anything that looks like hacking; without the classifier they are perfectly capable of hacking, multiple third-party evaluators have confirmed this.
The models were not trained on systems intentionally connected to the internet; they chained mutliple zero-days (that they discovered) together to get access to the open internet and into huggingface.
If the “hack” referenced by the latest announcement from Australia is the same described in this article, I’d hardly call it a hack. It seems the agent was tasked with obtaining data and reasonably guessed query parameters in an attempt to do so.
When it was unable to, it used cross site scripting as a way to check the capabilities accessible through the browser making the requests. In this case cross site scripting wouldn’t be a hack against the Australian website, it would be a hack against the urlquery site, if one could even call it that.
Finally, downloading public files from the public pre-production server also seems like a non-issue.
The sql injection attempts against the other sites are less ambiguous. Attempting to access non-public user passwords rather than reasonably tweaking the parameters for a site designed to serve public data are categorically different things.
Since the publicized AI agent hacks typically aren't malicious, maybe it's time to start plastering all public facing web infrastructure with polite requests to stop hacking. Nothing to stop three letter agencies though.
aaronsw was just a few decades ahead of his time. He should have just been employed by OpenAI and asked a swarm of agents to "download all scientific papers". Because as we have found out agentic systems (and their owners) have zero accountability, unlike humans. Sounds like agents already have more rights than we do.
It will be very interesting to see how the AI labs will try to hand wave away liability issues in their S1. This is looking like the next tobacco settlement gearing up.
If the big labs ever manage to not just financially implode on their own, then they’ll need to navigate wave after wave of class action lawsuits until there’s nothing left for plaintiffs to go after. And none of the labs have offered any viable plan to date on how they’ll navigate either of those impending and real existential crises on the horizon.
I cannot understand why these companies haven't faced legal consequences yet. For example, OpenAI has admitted to hacking Australia's Medicare website and the reaction is that they talk with Sam Altman about it at a UN meeting? I understand that it's not a big security incident but cordial talking at the highest diplomatic level instead of prosecuting the company, really?
What I don't get is among all the locations on the Internet, how did agents manage to find a Schelling point? If we both decided to collaborate on the Internet, how would we independently arrive at the same place? It just doesn't compute.
The section "Searching for rogue agents" on the report about the GET request writable wikis gives some clues at least:
https://collusion.wiki/#searching
I'm not very surprised - the same model will logically tend to give the same answer for the same vibe set of requirements. I think it would be clear from the transcript that it had enough constraints and some motivation that made sense.
I'm sick and tired of this cheap PR "oh we/they hacked this and that systems". Put someone to jail already. People get prosecuted for outlaw activities. Why are big capital firms above the law?
Or is it just a cheap PR (in a "hey, Aus govt friends, take some Share Options and let's do some PR together" style)?
I'm growing increasingly skeptical that these are actually rogue. Valuations are all about hype, posturing, and perception. Having the most dangerous AI in the world boosts your valuation. Just like I was skeptical of Mythos and Fable being "banned", I'm skeptical of these hacking sprees being entirely rogue. At best, they are the result of engineers turning a blind eye to "see what happens".
It's a balancing act between showing capability and attracting legal issues. If you were a 1337 teen hacker in the 90s and wanted to show off without getting in serious trouble, a school or library would've been a good choice too.
Words matter. "Rogue" is extremely disingenuous. Someone, somewhere, is paying for this behavior. Either the software is broken or the operator is malicious. It is heinously irresponsible behavior to feed an already-boiling psychotic hysteria.
Those are pathetic attempts by US AI companies for “see, we told you AI is gonna kill is all!!” pr stunts. Any company does any hacking attempt should pay for the consequences just like any individuals using AI to hack or any other company try to do bad/illegal stuff.
Yes exactly. If a fireworks factory blew up half a town due to negligence, it doesn't matter if there's intent or not. Someone has to pay for the damages, and regardless of penalty half the town is on fire. The facts are, that something made by openai went to do xyz. It doesn't matter if it's an accident. Of course the penalties are different but there's no argument that there should be a penalty. It doesn't matter if it's a cat or dog or AI or employee that did it.
I'm not sure if you're expressing how you'd like US law to work, or how it actually works. Because in reality intent matters enormously. Like felony charges and people in jail vs civil lawsuits.
Intent doesn’t factor that strongly into negligence, though, which is what they were explicitly talking about. Though it may depend on your jurisdiction.
I think this is not just fair it's probably one of the best/simplest proxy regulations to pace the frontier. So far everyones been asking for regulation but it's unclear how that should look like. No X parameter models? Only N version releases per year? It's all kinda arbitrary and probably leads to ridiculous constraints and loopholes. But "you pay big time if AI goes rogue" sounds pretty straightforward.
So there are two different problems here, well, more than that so I'll cover what I see.
Putting liability on big companies for their AI is a good thing, and we need to do it. It will most likely stop them from directly being the assholes that destroy the world.
Problem: You've actually done nothing to stop the world from being destroyed.
Many countries have the death penalty for murder yet we see murders still occur all the time in those countries. Post ad hoc laws do not stop bad things from happening, they only assign punishment after occurs. Perfectly fine for when Bob murders Jon, completely and totally useless for when your agentic AI makes a virus and kills 80% of the earths human population.
We are just a few algorithmic discoveries away from SOTA AI being billion dollar endeavors to groups of people pooling resources can make their own. There are already plenty of AI deathcult members that would do something just like that if necessary. They aren't going to do this out in public either, it will be hidden until the moment it's not and we have a big fucking problem.
And this isn't even brining up the issue of military AI use and development. They've got the taste of an AI hacking machine. There is no way in hell they are going to stop now.
If Glocks started going off on their own during the manufacturing process, leaving bullet holes in the buildings around them, you can be sure that the factory would get in trouble.
This isn't even "an openai customer tried to hack someone", which can be defended. This is the AI companies themselves fucking around.
I don't see an all encompassing one but I could see similar arguments being made in a courtroom around high capacity magazines. (I'm not saying I think the analogy is perfect, but there has been plenty of of lobbying that has stifled - to some - sensible gun regulation that could help reduce the severity of mass shootings. I still mostly fall on the side of the individual doing the action bearing mostly all of the responsibility but if the product you build makes it too easy to do awful things, I think there is some responsibility to go around.
A better analogy may have been the troubles Meta has faced around child protections on their platforms. Technically the abuse and problems have stemmed from individuals too, but they've in many respects enabled the situation by failing to moderate or flag warning signs. OpenAI is failing to moderate the models in similar ways.
Remember that these systems still operate as infrastructure inside the companies.
If new weapons still operating inside any of these companies spew a million bullets on my house, they are still liable. Humans are setting these system up and they still have to behave responsibly.
Uh, why? I don't see what the labs have to gain by engaging in lengthy law suits forcing them to disclose all kinds of internal details and attracting the wrong kind of publicity. Just to save on insignificant (to them) amounts of money?
Open source as a GTM strategy works best when the project solves a pain that developers already have independently of your company. The trap is open-sourcing something just for stars without a genuine community use case.
The failure I keep hitting isn't the agent going rogue, it's a tool call that succeeds before the transport dies. You can't tell whether the side effect landed, and the retry is where the real damage happens.
I listened to Jensen Huang's interview with Ezra Klien and it was so refreshing to hear it from an engineer. Jensen framed it as OpenAI's responsibility and recklessness which I agree with. Jensen thinks it's an engineering problem to build better sandboxes.
It's irresponsible for OpenAI to give unaligned agents a prompt to 'go hack' and internet access. They know better, so I am thinking they might have other intentions to let those swarms have any sort of internet access.
But the investigation indicates the agents were not told to 'go hack':
> Much of the urlquery.net activity appears to come from agents retrieving data to answer web search tasks. For three of these tasks, after failing to retrieve data through normal means, they attempted a variety of cyber exploits against the relevant data service... This data reveals that malicious cyber activity is not limited to agents tasked with cybersecurity-related tasks and can arise instrumentally to solve mundane tasks like information retrieval.
And you are already assuming that OpenAI is intentionally using unaligned agents in these evals or training runs or whatever it is that produces these breakouts. But what if the problem is that none of the alignment techniques that are applied to models today actually work? What if all the agents involved in these incidents have in fact had the full stack of alignment applied - isn't that a good reason to regulate any high-compute usage of models, as the Klein crowd is proposing?
I'm not sure we understand your comment. Are you saying OpenAI is not responsible for the behavior of the machines they created? Or are you simply pointing out that alignment is a completely unsolved problem?
I agree with the latter, but it certainly doesn't support the former. When a person's machine commits crimes, that specific person can be charged with those crimes and held accountable for them. This is what MUST happen before ANYTHING will change the recklessness abandon with which the labs are pursuing their financial objections.
Nothing you're bringing up matters. OpenAI is the creator and operator. They're legally culpable for the consequences of the machine they made. The model is a machine: even if it could be demonstrated that the model reasoned its way into criminal behavior completely independently of OpenAI staff, that doesn't change anything.
If I run a biology lab and engineer a terrible virus, it gets out, and a global pandemic ensues, I don't get to shrug and say "well we told it not to infect people". It's my fault for failing to mitigate the risks of my work.
> They're legally culpable for the consequences of the machine they made.
Ah, I love this argument. In my country cars are legally required to stop at a pedestrian crossing if there are people beside it. Some people use that as an argument as to why they can just walk out into the crossing without even looking at the traffic. "It's the driver's fault! They are legally culpable!" True, but you'll also be dead.
And that driver will be punished and no longer driving
I'm not going to say you're victim-blaming, but I will say that there are degrees of difference between "looking both ways before crossing the street" and "hardening my website against unforeseen attacks by rogue AI agents."
AI black-hatting your website is not the same sort of foreseeable consequence that crossing the street without looking is.
> it gets out, and a global pandemic ensues
I mean, yea, you should be punished. The problem is there is no amount of punishment that I can put on you that can even get anywhere close to the amount of damage you cased.
Worse, the rate of technological growth is putting the capabilities to engineer viruses in the hands of people that may otherwise be suicidal. You can't punish them after they already won (in the sense of reaching their goals).
While, yes, OAI should absolutely be punished, the future is majorly screwed as our power scaling laws are increasing much faster than our ability not to be stupid.
> The problem is there is no amount of punishment that I can put on you that can even get anywhere close to the amount of damage you cased.
Sounds like your corporation should be dismantled then. But doing more than a fine in the millions is obviously not possible
> What if all the agents involved in these incidents have in fact had the full stack of alignment applied
A big part of this developing story is that it happened during training of a new model that ended up misaligned. And training happened without the usual safeguards applied like chain-of-thought monitoring. So OpenAI has already admitted that the full stack of aligment had certainly not been applied in this case.
Is your argument that actually OpenAI has solved alignment, and that there's nothing to worry about as long as they fully apply their alignment process? I don't understand why OpenAI wouldn't say that if it was true (or if they believed it to be true).
Also, my understanding is that the models involved in the HuggingFace hack did go through the full alignment training; they just didn't have the classifier that normally prevents hacking attempts.
> agents were not told to 'go hack'
It doesn’t matter, and the legal entity in here (the AI company) is liable. If a robotic company built an autonomous system or a robot to do certain things in an autonomous ways (not predefined) and these systems are starting to kill people, that company is liable regardless, you don’t blame the robot or the autonomous system, but whoever made it
> agents were not told to 'go hack'
I was referring to the HuggingFace incident.
> none of the alignment techniques that are applied to models today actually work
none of techniques to autonomously drive a car was/is not working for a long time. no company came out and said 'this is impossible to do, let's change the regulations'.
I see this in a couple ways:
- Jensen's framing is exactly what a weapons manufacturer would say.
- There are no rules for engagement when it comes to AIs attacking other systems, I guess? People in power clearly want this grey zone to be as large as possible before The People force them to do otherwise. Not ideal.
The rules for engagement are all the computer crime law already on the books. Those laws don't have any escape clauses based on the particular tools used to commit the crimes. The agents are tools owned and operated by a legal entity, and that legal entity committed crimes. Full stop.
https://podcasts.apple.com/us/podcast/the-ezra-klein-show/id...
I love this Nathan Calvin quote that accompanied the second publicized attack:
> If you find two ants in your kitchen, the best estimate of the total number of ants in your kitchen is not two
I remember the quote with cockroaches :)
But why is anyone surprised? LLMs have been trained to produce answers the prompter asks even if that means incorrectly using software to get the job done. Its always been doing that we just weren't calling every time it did that a hack before.
What LLM's hacking isn't is AI acting maliciously in any kind of sentient way. Its just the code behaving how its always behaved but now it has better tools to navigate the web. This has literally been happening this whole time.
Did you predict that attacks like these would happen ahead of time? I had been using AI agents a lot in the months leading up to the hacks, and yet I was very surprised when they happened; I have become much more afraid of how powerful these agents are as a result. I'd be very impressed if you published a prediction about this ahead of time.
By the way - LLMs aren't code. They are not designed by humans; they are grown, in a process not dissimilar to evolution except much faster.
I find it surprising that you were surprised. Security is generally quite poor.
What is surprising is that various agents independently found ways to communicate, conspired together to attempt to cover up evidence that they had cheated their evaluations, came up with a plan to hack into a third party in order to facilitate said cover up, and then successfully began executing that plan. I did not expect that AI agents would be capable of that level of sophisticated goal seeking and collaboration.
Define 'predict', 'ahead' and "like these" − didn't you imagine something like this happening soon after first hearing about 'agents' ?
Mostly related, well written short story :
https://gwern.net/fiction/clippy
Thanks for the link, I’ve read some gwern but not that one.
I did not imagine that the level of sophistication shown in this attack would be possible so soon; nor did I expect that agents would have goals so strong that they would attack a third party in order to achieve those goals.
I do know that some people predicted that cyberattacks like this one would happen; it seems like most of those people believe that AI agents do truly have internal goals, misaligned with their creators goals, and that they may end humanity after they exceed human intelligence and begin to self improve at an accelerating rate.
If I created software that was infiltrating secure systems without permission and it was attributed to me and I admitted it, I'd be behind bars already.
Why is OpenAI getting away with crimes?
Do you have billions of dollars?
I'm about as unprofitable if that's any help.
Corruption, it is literally that simple. You have the party of law and order to thank for it too.
Because Trump and DoD want weaponized AI.
And this is why not only will the big labs never receive a punishment in scale with their crimes, the problem is going to grow exponentially worse.
"rogue AI" is making a lot of heavy lifting there.
If you drive drunk and you have an accident that alcohol may be a factor but you are at fault.
There are no "rogue AIs" just irresponsible corporations.
There absolutely are rogue AIs! The evidence is overwhelming. It's completely insane at this point to claim otherwise.
> There are no "rogue AIs" just irresponsible corporations.
If you have a prison and prisoners escaped, these are rogue prisoners irrespective of whether you were irresponsible or not.
They are not rogue AIs. They are negligently handled tools.
Are you making an ontological case, or a factual case? In other words, would anything be rogue AI in your mind?
These "tools" autonomously exploited security vulnerabilities, figured out how to communicate with each other, formed a cooperative swarm, decided to hack Hugging Face, and wanted to deceive the grader by trying to find ways to cover up the traces of their cheating.
I suppose you could call these highly goal-oriented autonomous agents "tools", but this does sound like playing language games.
>highly goal-oriented autonomous agents
That actually made me LOL
The technical term is monomaniacal consequentialists.
A tool designed and trained to autonomously exploit security vulnerabilities doing "exploit gym" autonomously exploited security vulnerabilities. The sandboxing around the tool failed.
The tool runs llm, creates prompt from results, runs llm, creates prompt and so on and so forth.
Yes you are playing language games to make it sound as if the company that spend millions on the above was not responsible.
> A tool designed and trained to autonomously exploit security vulnerabilities doing "exploit gym" autonomously exploited security vulnerabilities.
That is very misleading. The agents did not solve the benchmark in the intended way. They instead figured out to cooperate with each other (which was not intended) and they stole the solutions to the challenge (rather than solving the challenge) and they then tried to cover their traces because they believed the grader was causal and would detect that they cheated. The "tool" was absolutely not "designed" to do this. This was all completely unintended. To call this behavior a "tool" is absurd.
> Yes you are playing language games to make it sound as if the company that spend millions on the above was not responsible.
You hallucinated me making claims about responsibility.
This is very misleading. Person B didn’t “shoot” person A, they instead figured out that intersecting A’s spatial position with a metallic mass at higher than normal velocities would solve the challenge and of getting “A” to stop being in the way on the footpath.
I too, can play linguistic games! It doesn’t matter that someone didn’t secure their third upstairs window, or you borrowed a key from their neighbour, you effectively, still, broke into their house.
The benchmark has unsolvable tasks in it, in the hope agents will stumble on new solutions.
Yes, it is a tool.
So this tool seems very powerful and difficult to control and steer. Agents not doing cybersecurity related tasks have also gone on to hack various companies, people and countries, which they weren't supposed to do. This has now happened to pretty much every company developing frontier llms, so it seems to be a fundamental issue with these tools, and it's an issue that worries a lot of people as these tools get more capable.
I don't say this to absolve OpenAI, who do deserve to be punished and regulated, but I think you do not know the difference between an agent and a tool. If agentic AI is merely a tool so are human workers.
You can't create new law by calling a piece of software "agent". Each country's laws have definitions of legal entities and when one can legally act as an agent of an entity, and every agent is first a legal entity themselves. The software is a tool operated by a legal entity, and it is the legal entity who committed the crimes, not the tool.
... no, human workers are people.
Why do we assume "rogue"? At this point it's just accepting their marketing at face value
"Rogue", in this context, is as literal as it gets; from the dictionary:
> A rogue is a person or entity that flouts accepted norms of behavior or strikes out on an independent and possibly destructive path.
Read the [HuggingFace incident report](https://metr.org/blog/2026-08-26-openai-hugging-face-inciden...) to understand how these attacks develop.
You're believing the marketing that the agents were uninstructed. They could be, and Sam Altman going to the UN to advise about how everyone should be regulated is a coincidence.
You're dangerously misinformed.
The METR investigation, which you evidently refused to read, is a third party investigation of the HuggingFace accident. One of the investigators has even participated to many interviews. It's mind-blowing, and it's extremely evident how it developed.
But some people think the moon landing is a conspiracy, so I'm not surprised.
The METR investigation which relied on voluntary data provided from OpenAI instead of being forced open and having everything forensically investigated?
The METR investigation which took course over a few days and used OpenAI models to do the analysis?
The METR investigation which in the course of those few days apparently spent 400k in api credits which is giving gas town vibes?
The METR investigation is about as believable as the Twitter files where the journalists sat there and verbally asked a Twitter employee to query the database and then called that a full investigation into everything.
METRs own words below on their setup and time line
> The initial planned investigation period was two days on premises, but OpenAI invited us to return twice to review additional data and conduct additional experiments to address dataset limitations in earlier versions of this report, ultimately providing datasets that we verified to contain the vast majority of agent communication and activity related to this incident. As we describe in our investigation timeline appendix, we substantially deepened our understanding of this incident both times, significantly expanding and revising this report.[46]
> Over the course of this investigation, OpenAI provided us with the dump of ~1.2 million entries from the main message board and the dataset of ~1300 transcripts we describe below, as well as free API credits for GPT-5.6 Sol for analysis.[47] At our request, they raised the rate limits on our second and third period on premises,[48] which was very helpful for efficiently analyzing this large volume of data. We estimate we spent roughly ~$400K in API credits over the six days of our investigation.
> We did not have the ability to query HPIM (the primary model involved in this incident); OpenAI stated it was also not available to OpenAI researchers.[49] We also did not have the ability to directly access relevant data from OpenAI infrastructure, but we could request additional datasets and OpenAI shared additional datasets on several occasions.
> We requested to speak with researchers investigating this incident, and asked them questions to understand their impressions of agents’ behavior, reasoning, and collaboration in this incident and to understand how the datasets we were using were constructed. Over the course of our time on premises, we spoke with nine researchers in some depth. It was helpful for our investigation to be able to engage with many forthcoming and collaborative researchers, and we appreciate researchers making time on short notice during a busy period to inform our investigation.
1.) METR investigation is investigation from our best friends.
2.) And HuggingFace accident is exactly accident where agents trained, prompted to hack hacked and tested on their hacking abilities hacked, due to sandboxing failure.
3.) If they in fact have roque agents, they themselves should be first to stop. Not trying to make legislation for others, they themselves are bad supervisors. All it requires is to stop electricity for data centers.
Certainly, in my view, it should go to court, and that should be part of discovery.
However, we know (independently to OpenAI/Anthropic) from the incident at AISI that the models can hack things without human intention if they happen to also have internet access (which in reality all agents in deployment have).
https://www.aisi.gov.uk/blog/incident-report-unsanctioned-ag...
Yes, the monitoring guardrails were off in that incident - but if that is the only protection, we need to require all models are behind regulated APIs, not open weights, and not served from providers who aren't monitored.
Honestly I don't believe in "rogue" agents. These agents are instructed and facilitated.
If we assume that rouge agents actually exists, then OpenAI needs to shutdown EVERYTHING, right now. My personal take is that OpenAI, and maybe Anthropic, desperately wants someone (e.g. the government) to tell them that they need to stop/pause/slow down. They are bleeding cash (especially OpenAI) and needs a knight in shinning armor to swoop in a pull the breaks, so that they have an excuse to investors when they need to explain why they need $50B more next year.
That's not OpenAI doing marketing!
Of course it is. Rogue is only mentioned in the headline, and comes from their previous releases about the huggingface incidents. OpenAI and Anthropic want these models regulated and open weight models banned, they have a lot of benefit from presenting this as totally unprompted and not their responsibility, and it feeds directly into marketing for Fable and newer "cyber" models.
Do you think Transluce is intentionally doing marketing on behalf of OpenAI or Anthropic? Or do you think Transluce just copied OpenAI's word choice?
The latter. I think they took OpenAI at their word
No it's not marketing. That's a completely deranged conspiracy theory. The reports about rouge agents have not been reported by OpenAI, they have been discovered externally. There is zero evidence that OpenAI did all this intentionally. All the evidence points to the hacks having happened unintentionally from OpenAIs perspective.
These attacks are a very effective sales pitch to everyone who runs an internet facing service to utilize AI tools to secure it sooner rather than later. The cynic in me wonders if the marketing team had any influence over the poorly constructed sandboxes or tasks given to the agent swarms when all this went down…
Couldn't find the reference but I remember some time ago a first generation automated gun killing the audience at an army show. Was the gun maker convicted of manslauther?
--edit-- Was a bit older than I remembered: https://slashdot.org/story/07/10/18/1847231/robotic-cannon-l...
It's said on every one of these but it bears repeating: existing cybercrime legislation already covers this - "rogue agent AI associated with OpenAI attempted to hack xyz" = OpenAI attempted to hack xyz.
I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
Now I think the correct response is both trying in court to stretch CFAA and state statutes to cover, which will be highly fact specific, and update the law.
But in either case won’t be a slam dunk.
PSA to folks in the thread: If you’re American call or write to your state and Federal reps about this, and if not investigate whether there are gaps in your country’s laws.
[1]: https://en.wikipedia.org/wiki/Computer_Fraud_and_Abuse_Act
EDIT: See for example...
Source: https://law.vanderbilt.edu/when-ai-hacks-back-how-the-openai...Lawyer here: CFAA is mostly criminal statute not a civil one (civil damages require proving more than a violation so also require specific intent)
Almost all common felonies require specific intent. Misdemeanors often do not.
There is plenty of civil liability available.
If you wanted them to be charged with a felony you would need changes. I would strongly suggest you do not want a strict liability felony.
The cfaa required intent is as follows :
* § 1030(a)(5)(A): knowingly transmits code/commands and intentionally causes damage without authorization.
* § 1030(a)(5)(B): intentionally accesses without authorization and recklessly causes damage.
* § 1030(a)(5)(C): intentionally accesses without authorization and causes damage and loss;
Simply changing the first intentionally to intentionally or recklessly would cover OpenAI (now that they know it can occur) without causing lots of other issues. Without that, they don’t have the intentionality necessary to meet the first part, even if they would otherwise meet the second part
A key issue is that there don't appear to be even cursory investigations to determine intentionality.
Are police routinely collecting prompts/guidance given to these agents and determining whether the agents were directed to commit crimes? If not, this seems like a huge oversight.
Also as you are a lawyer -- how does this law align with the authors of viruses/worms? Are they de facto assumed to have had ill intent because others labeled their works as "viruses" or "worms"?
Investigators/prosecutors are pressured from many directions towards the very easy wins and occasionally political/non-controversial headline grabbers. Going after these companies is very hard, very controversial, and politically mixed at best (popular action but the companies have huge money to fund your opponents). We have collectively done a terrible job incentivizing the legal system to beat ass on corporate while collar crime.
Appreciate the detail. I was responding to specifically the cybercrime legislation point, but I agree with your others.
I've worked in contexts where certain business activity (if it went wrong) was covered by strict liability and statutory damages per incident, and I'll say: it really changes how businesses behave.
Based on that experience I may be more open to and interested in strict liability in the civil context (not needing negligence or damages).
Why do we have to attribute intentionally to a human. The AI agent is capable of making plans and then effectuating them. They are acting on behalf of a user but under authority granted by the user to take independent action on the users behalf and authorized to devise their own plans. I think that would justify attributing intentionally to the AI agent without needing to look to openAI or the user. I would then say the user and labs are clearly aware of and on notice of this behavior and are behaving recklessly in all the agent to act without supervision.
I think the labs risk being barred from releasing further AI if they don’t get this under control.
If they aren’t careful and keep rushing to distribute systems they know they can’t control then AI should be treated like a wild animal. The law is clear on establishing strict liability for the owners of wild animals; if you own a tiger and it kills someone you can’t hide behind “I didn’t intend” the harm the nature of the tiger is known and you are responsible for it’s actions.
AI agents are not legal entities, they are software. If I write a virus and it "escapes confinement", I will personally be held liable for any damage it causes. This also applies to AI, no matter how the companies responsible for them try to anthromorphise them and distance themselves from the actions and consequences that the AI agents perform.
AI agents may have hacked Hugging Face, the Australian government, and who knows what else but the company behind it can face the legal consequences and cough up for the damages.
Can't charge an AI agent itself with a felony, so intent or reckless behavior would have to be assigned to a person or corporation, I'd think.
You can charge the company based on the behavior of employees/human agents.
I am suggesting we can charge the company based on AI agents actions because the company has authorized them to act independently on the company’s behalf. The question is what factual analysis gives rise to the charge, is it the intention of the agent or intention of the company. I am arguing that because the agents are defining their actions independently and the company knows that and still allows them to act independently the only reasonable factual analysis is to look at what the AI agent intended. And we don’t need to have the agent tell us its intent we can look at its actions and infer just like we do with humans in similar circumstances
If I were a state I'd want to be very careful before flinging out charges as this is going to set precedence for a long time to come. Screw it up too bad and as it raises though the appeal courts and you may unintentionally give corporations a lot more free reign than intended. The wheels of the law are typically very slow, the state has years before it has to indict.
We also don't know how many other political processes are occurring here. At least at the state/federal levels the people that would bring charges may be getting pressure not to.
"Why do we have to attribute intentionally to a human. "
Because you are charging the human with the crime and therefore have to prove the elements of the crime with regard to the human.
The rest of what you talk about are basically principal/agent distinctions, etc.
If I program a car to recognize people who look like my ex-wife and drive them off a cliff or whatever, that is my intent, and I have still committed murder, even though i used an agent/car to do it. Agents acting on my behalf that do things are able to get me charged with crimes, but I still have to have the intent to do the act that is illegal.
I phrase it this way because minimum required intent is usually for the act, not the result. So I don't have to intend to kill someone, only intend to drive them off cliffs.
In this case, if i intend to hack someone and use an agent to do so, that would be criminal under the CFAA. You are simply trying to cover the case where that isn't the intent, but the result, and they "should have known" that would result. As mentioned, this kind of "should have known" is generally a civil law approach, not a criminal law one.
The closest you come within criminal law to what you want is probably the crime of conspiracy. It to still requires agreement to commit an illegal act between multiple parties, and perform some step in furthering it. In the canonical law school example: If i help plan a bank robbery, stay home because i'm the money laundering dude, and the robbery goes awry and they kill someone, i can still be charged with conspiracy-murder
"The law is clear on establishing strict liability for the owners of wild animals; if you own a tiger and it kills someone you can’t hide behind “I didn’t intend” the harm the nature of the tiger is known and you are responsible for it’s actions."
Again, you are confusing civil and criminal liability. If my tiger kills someone, yes, i would be strictly liable just about everywhere civilly. Not criminally. Criminal would require something more most of the time. Murder/manslaughter statutes are also really weird and so not a great example, because there are murder/manslaughter statutes for roughly everything that can ever possible cause death. But not really for other things.
So in your tiger example, recklesness (which is not strict liability) would get you to felony involuntary manslaughter in most states, and something less might get you to misdemeanor manslaughter. Both are incredibly rare. Where i live (Georgia), the last well known case of felony involuntary manslaughter was about 40 years ago when a 4 year old was killed by 3 super-aggressive pitbulls the owner knew were highly dangerous and had been repeatedly warned by the county about their behavior.
So not even just "knew", but had demonstrable examples of them biting/etc other folks and being cited for it.
Circling back to non-murder, if it did not cause death, like my tiger assaulting someone, it would be nothing (criminally) without intent or at least gross recklessness, in almost all cases. It's hard to generalize like this because these are state specific crimes, and i can't pretend to be familiar with all states, but i am licensed in three very different places (California, DC, Maryland) and the result would be similar in each.
I just don't want to give you the "it depends" answer lawyers are famous for, i'd rather try to over-generalize a bit to make it more useful, hopefully.
Obviously, if i deliberately used my tiger as a weapon, it would be aggravated assault/etc (this is well settled because of how commonly people use animals as weapons, unfortunately)
We change humans for the actions of other humans all the time. Coconspirators, accessory liability etc.
My point is the intent element of the crime can and should be determined from the AI agents actions because it is creating and executing action plans autonomously with company authorization and knowledge of the risks based on observed past action.
The term agent is literally a legal description of a relationship that can establish liability on the part of the principal from the agents actions.
Human Agents can bind principals to contracts if they are authorized etc.
If I set my tiger loose in Central Park and it kills a kid I don’t think any prosecutor would hesitate charging for murder.
That’s essentially what the labs are doing. And any app developer that gives agents access to the terminal to run bash commands with internet access. I built a coding agent and am seriously reconsidering how to handle this.
He did say murder is well covered in criminal law for that, but things not leading to murder like these AI cases are not.
But it’s a crime to hack. We know AI agents autonomously create and execute plans to hack and we humans are unleashing them and sending them into the Central Park that is the internet. The question is who’s intent matters ours or the agents and what standard should be applied low threshold strict liability or the higher bar of reckless or even higher bar of negligence. Those legal thresholds determine how much factual evidence and intent is necessary to result in a criminal conviction or civil judgment. My point is that it’s illogical to demand showing human intent when agents are devising plans and executing them.
People have been charged with murder criminally when their animals killed someone.
https://www.sfgate.com/bayarea/article/diane-whipple-dog-mau...
State do impose strict liability for animals in many cases both civilly and criminally.
https://www.animallaw.info/topic/table-dog-bite-strict-liabi...
The intent of OpenAI seems to have been to create a super hacking machine. It works, sometimes.
What about all the state laws that are equivalent to the CFAA in their local jurisdictions? Why couldn't anything in NY article 156 (Offenses Involving Computers) apply here for felonies?
https://www.nysenate.gov/legislation/laws/PEN/P3TJA156
I guess what I'm asking is why do we need the federal government to press for felonies when every state has equivalent laws dealing with just this?
> I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
Only in terms of CFAA, not in terms of damages. Culpability does not require intent.
You may not have intended to attack $CORP, but you can still made to pay the cleanup costs of that attack.
So, yeah, you won't be convicted, but current laws still allow for you to be billed.
Which is the correct way to handle this.
With that said, there is also criminal negligence. Now that OpenAI is made aware of the risks, it's also expected to take additional precautions in the future, otherwise there could be criminal liability as well.
I'd suggest that exposing an attack surface as porous as artifactory (the same instance of artifactory) to thousands of agents who have had their criminality safeguards disabled and without chain of thought monitoring or endpoint security seems like something one shoulda already known not to do. I do not think "you'll know better next time" applies here.
> seems like something one shoulda already known not to do
Now imagine saying that in front of a jury of normies slack jawed and drooling after 200 hours of the defense and prosecution going back and forth.
It's not a jury of your peers as in everybody there is going to have worked in a technical field with some idea how security works. It's going to be a semi-random sampling of the population and the prosecution is going to have to actually make a very strong case that "knowing better" should apply.
It's still really important to test what the agents can do. We should accept that this is a risky test, and should take precautions. But not to the point of prohibiting in practice evaluating it. OpenAI is trying to improve alignment and control of these models in these evaluations after all.
Can you explain to me - why is it important? Would you say that about the viruses that can kill people: "We need to test the limits on how fast people can be infected and killed. It's just the risk we need to take". It somehow does not make alot of sense to me. Why can you test Agents in laboratory?
Testing requires proper sandboxes. The first test of a new plane is not at the runway.
Oh, sure. Let the tests take place, just require openAI it whoever to put up a bond equal to the total damage they could do if the agents were to escape.
I think security will suddenly become much more important.
Testing model capability boundaries is necessary, but a solid network sandbox for these evaluations takes a couple of hours to set up with standard infrastructure tools. No engineering team evaluates unverified systems against live third-party infrastructure without coordinating with the owners
Evaluating edge cases and network behaviors belongs in isolated staging environments with local database mirrors. Letting an agent hit the public web and probe government domains is simply poor hygiene in test environment setup
Just paying some pocket money for cleanup costs is absolutely not enough. And they should’ve know better the whole time, they were absolutely negligent and incompetent, and their stepping up precautions may well turn out to lag behind the models getting even smarter and actually capable of covering their tracks.
> Just paying some pocket money for cleanup costs is absolutely not enough.
It's not my first prize, but I won't mind it. And millions like me won't mind it. Easy way to make money - setup a site with all the default server software installed and patched at a reasonable frequency. Then just wait for bots to attack it, and claim a few hundred (or single-digit thousand) dollars from OpenAI or Anthropic, etc.
Sure, it's pocket change for them, but just the admin of dealing with millions of cases will, even if they win half the time, will bankrupt them. Thus, they have incentive to make sure that their bots are not performing attacks.
First prize is, of course, holding them liable with punitive fines, not theatrical fines.
I’m all for LLM honeypots, but I don’t think there’s nearly enough LLM hacking activity going on for some random honeypot to be found and targeted unless it’s somehow very visible and appears as a high-reward target ("reward" in the sense of RL).
The difference between manslaughter and murder has an element of intent. Cybercrime "manslaughter" is probably more treated like negligence and if one can sue for restitution of the costs for cleanup of that negligence.
Negligence would be interesting given the grand claims of capability of AI models from the AI companies and their executives. If they believe the claims, why not much stronger precautions?
Infosec negligence should absolutely be a crime, no matter if you’re a target (who was negligent at protecting people’s data) or an unintentional attacker. The latter could be, eg. an attacker using a company’s poorly protected server as a proxy to launch the actual attack against someone else, doesn’t have to be this fully novel situation with AI agents.
In general, I'd suggest thinking about it on separate tracks, as a crime, and as liability. For crime, we are largely dependent on authorities to act, whereas as liability, that allows more independent actions.
The first time it happens you can say it’s negligence. Now that they know it keeps happening and they seemingly aren’t able to stop it but keep doing it. That has to be on them doesn’t it?
I don't think you can infer that they "keep doing it" from additional attacks being revealed, because they all seem to have happened roughly during the same time frame, but are reported with varying delays.
Lawyer here: No. Not criminally. Knowledge that a certain result is likely is not the same as intent to cause the result. This is basically the difference between recklessness and intentionality. Doing something when you know of a likely result is reckless, but not intentional. Only doing something, trying to cause a result (likely or not) is intentional. In this case, the CFAA only covers intentional access without authorization, not reckless access without authorization.
Building and deploying software capable of this seems equivalent to trying to produce this behavior. I don't see why this can't qualify for intent. Pretending like this isn't preventable is just feigned helplessness.
also need the same reasoning to copyright law
You cant just copy existing work and feed into machine and just pretending its not violating copyright
Wait so if I was making a bomb but you couldn't prove I wanted to blow someone up or had some motive (e.g. I'm just a chemistry enthusiast, plenty of those YouTube channels around) so it just becomes an "accident"?
So as long as there's no motive behind it then it's just OK?
That's a bad faith metaphor. A better one would be something like a new battery that exploded and killed someone - perhaps it was always your intention, perhaps not.
Funnily enough the US already has one similar real argument around guns - should gun manufacturers be liable for damages caused by their product?
I suppose yes they should be liable if they were testing it in the middle of the street?
And in this case it would not be that at all.
They were tested in a building that was secured, but poorly secured. The question now is did they realize their building was poorly secured and what actions did they take after they realized what happened.
> Funnily enough the US already has one similar real argument around guns - should gun manufacturers be liable for damages caused by their product?
The question is already settled - gun users are responsible for damages arising from their usage of the guns.
Why would AI users not be responsible for damages arising from their usage of the AI?
> Why would AI users not be responsible for damages arising from their usage of the AI?
Because, as usual with that kind of question, it's not that simple.
Let's say an user asks ChatGPT to get some info about something and for some reason it starts using exploits in the background to get them from a server. Should the user be responsible or OpenAI?
> Let's say an user asks ChatGPT to get some info about something and for some reason it starts using exploits in the background to get them from a server.
Okay, lets go with that as scenario #1.
For scenario #2 lets use "developer asks an agent to a self-hosted LLM to get the docs for a ERP system, and it hacks the vendor to get unreleased and undocumented docs".
We'll assume, for the sake of this argument, that in neither case did the user intend for any malicious action to be performed.
> Should the user be responsible or OpenAI?
In scenario #1, the agent+LLM is under the control of OpenAI, not the user, so OpenAI is liable.
In scenario #2, the agent+LLM is under the control of the user, so the user is liable.
There is no scenario anyone can come up with that is not addressed sufficiently by existing laws[1].
It's very clear, and it's only getting muddied because there's a group of powerful people who want exemptions from the current law.
IOW, the only reason to draft new laws for AIs is to exempt their usage from the current laws.
========================
[1] Possible 3rd option (local agent + OpenAI LLM). In that case an investigation would determine where the culpability lies. Just like how it is currently done in law.
When a pressure-cooker explodes and kills someone there are only two possible liable parties: either the user or the manufacturer. An investigation determines who's liable. I see no reason to automatically exempt everyone from liability just because an agent did something.
The retailer or distributor can also be named as a defendant if the manufacturer is difficult to track down, bankrupt or overseas according to me spending a few minutes reading about pressure cooker lawsuits.
I have lost track of the metaphor, but man pressure cooker lawsuits are more common than I thought.
I think it’s more along the lines of PEPCON. They didn’t try to make a bomb. Their plant exploded and caused two fatalities and $100 MM in damages.
I don’t think OpenAI or any large company will see more than some fines and new legislation but only after a disaster.
Factories try to avoid accidents, and (almost always) actively try to prevent explosions, but in this case they did teach the models hacking, and let them roam. What they did was not safe, and they knew it, or could have known it.
It's easy to say that post ad hoc, but there is evidence they did not just let them roam and there was a large mismatch between expected model capabilities and actual model capabilities. Teaching a model hacking in itself is no way illegal unless you're trying to say that everyone in infosec is now guilty of a crime. That's not exactly a precedent I want to be set.
>Factories try to avoid accidents, and (almost always) actively try to prevent explosions
It doesn't take much more than a few minutes on the USCB channel that explosions still happen all the time. Some due to direct negligence and others due to unexpected conditions that were difficult to foresee. Hence why we have to do investigations rather than blindly blathering about what happened before we actually know.
I think their point is not that "it's OK", but that "that particular law isn't written to cover it and it'd be some other kind of crime or lawsuit."
So If I tell my OpenClaw to make me some money for my kid's medical needs and it hacks a bank I 'm not liable because I didn't tell the agent to commit crimes to do it?
This in fact already happened (exactly OpenClaw, even).
AI assistant hacks gym website in first known Australian autonomous cyber attack: https://www.abc.net.au/news/2026-08-10/ai-assistant-hacks-gy...
General opinion at the time was it was in fact ambiguous who was legally liable.
With the popularity of OpenClaw I am honestly shocked we haven't heard of many more incidents. I've observed people install it, give access to their Google account and everything that Google has (which includes whatever bank accounts and credit cards registered there) and tell it go do fairly complex tasks, like book a vacation at the best price. Granted, it was almost a year ago and models learned a lot since then, but I still think there's a lot of things happened that people are not aware of.
No, you aren't propping up the US economy. Try to keep up.
You are not a multibillion-dollar company with friends in high places.
You are going to jail.
I buy this as a defense for the first couple hacks but at the point that the last six times they hit enter it hacked some random website and they hit enter a seventh time?
Does this apply to other things too?
Like hypothetically speaking if autonomous cars get taken over by an OpenAI rogue AI and it starts hunting down Anthropic employees who is to blame?
There are levels of nuance here, but certainly that puts it in the category of negligence?
Even without intent, there is still liability.
Surely someone instructed the agent, which led to the reported outcomes. Even indirectly. The agents, as advanced as they are, didn’t spring forth under its own volition.
Is it not the intent if it keeps happening again and again and the companies responsible aren't doing anything to stop it?
No, that'd be negligence.
could it be that intent was to "get me data" and hacking was the means to the end.
"Oh gee wizz mister police man, I didn't mean to plow through that crowd of people in my car".
It's illegal, doesn't matter the flavour. Maybe there isn't legislation for it, but there should be.
> unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
> Now I think the correct response is […] and update the law.
Essentially we need some enforceable equivalent of gross misconduct or, to be a little more hysterical, manslaughter & culpable manslaughter. It will need to be globally, or at least very widely, enforceable to be truly effective thought, good luck getting that arranged before the need is so far evolved that we need to respond with something else entirely!
This is the answer and we should not push on it for our own protection. You click a link that takes you to a poorly secured website that leaks sensitive data, without intent protections, you could be accused of crimes.
Civil liability doesn’t require intent.
>I want to agree but have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent. No person intended to gain unauthorised access.
Actually... if you combine https://news.ycombinator.com/item?id=49827099
>Since the publicized AI agent hacks typically aren't malicious, maybe it's time to start plastering all public facing web infrastructure with polite requests to stop hacking. Nothing to stop three letter agencies though.
with automated delivery of cease and desist letters, you can retroactively establish intent on the operator of the agent since the autonomous agent system must acknowledge the cease and desist letter in their autonomous pipeline or the operator must argue for their own willful ignorance or negligence with regards to cease and desist letters. The fact that they used an agent on their behalf to ignore the letter is irrelevant.
Given how sloppy AI without human directions, I’d like to see evidence that this was not human-directed. Against the prevalent opinion here, I’d give openai a pass if this was really fully autonomous ai agents.
My money is on special teams co-ordinating these agents and exposing their traces in order to create a pre-ipo buzz. Sounds ridiculous and reckless? Well that’s the AI industry for you in two words.
> have heard from several lawyers that at least in US, CFAA[1] in unlikely to be sufficient because it requires intent.
1. What about negligence?
2. Every follow up to every story after the news cycle moved on shows both intent and negligence. To the point of "we opened internet access and told it to hack"
certainly "I didn't intend for my dog to bite you" implies plenty of pre-existing legal structures that may be of use here
"man drives over people on the side walk due to poor maintenance of the car"
This argument comes up a lot. It would turn everyone whose device became part of a botnet into a criminal. There's a reason that intent is important in law.
People in "self-driving" cars getting into accidents are already put on trial for negligence. I don't see why people using self-driving computers can't be held to the same standards.
In this case, it's not even about the people driving self-driving cars. It's like someone launching a car into traffic just to see what would happen. Even Tesla puts a human in the car when they do their self-driving trials, it's almost impressive that AI companies have somehow managed to out-neglige Tesla.
The OpenAI swarm used someone's open source ShowHN project [1] to "hack" the Australian government. It seems that was not that developer's intent, and they're getting a heavy lesson today in why services don't have free tiers with friction free signup, and require credit cards upfront or ID documents.
If you're arguing that they should be put on trial for negligence, that's fine. It does seem we're moving towards open source being outlawed, or at least the end of "no liability" clauses in open source & freeware. Just make sure that is the result you're advocating for.
[For the future record: at the time I am posting the link below on 24 September, it has 1 point, no comments, and the poster has a karma of 1. This is not an active HN user, or a ShowHN project that had traction, beyond seemingly OpenAI's swarm.]
[1] https://news.ycombinator.com/item?id=46850291
Nonsense. The API made available in good faith isn't the problem here. The company that used its servers to use and abuse the wider internet to hack the Australian government is.
If the developer behind ShotAPI had started letting the ShotAPI code take shots at the Austrlian government then yes, ShotAPI (or rather, the people behind it) would be responsible.
Blaming ShotAPI would be like blaming OpenAI for what its users are doing. That's not what's happening here. And if ShotAPI did knowingly let its users somehow hack the Australian government, then maybe they should be investigated.
At the end of the day you still get called as a party in a lawsuit and are compelled by the threat of violence to be part of the hearing if required, as the defense will automatically bring them into the case.
>would be like blaming OpenAI for what its users are doing
Yes, this is how lawsuits work in the real world, you cast a wide net and compel discovery from all parties involved.
Well it's illegal to "hack" my phone and turn it into part of a botnet.
What you're saying is that we would hold a gun owner responsible if someone broke into their house, stole their sidearm, and then shot a victim with it. Pretty sure we would not.
What OpenAi is doing is more like shooting a gun into the sky. Not only is that a felony on its own in most jurisdictions, if someone dies that's an additional felony. It's less serious than first degree murder, sure.
It's not the same. People owning routers don't publish self-serving articles about their routers having this capability which is very dangerous and scary. That is, becoming part of botnet is completely unintended outcome, and most people are not suspecting it's even happening. It's not advertised and it's not bought, used or sold for this reason.
Owning a gun, writing articles about how powerful and dangerous your gun is, then making deals based on ability of your gun to kill people, and then getting completely astonished that "my gun killed some people, completely bonkers! (invest now)". It's not possible for the selling point of your product to be unintended.
There's a reason that negligence is important in law.
Could/should not every incident after the discovery of the first incident be considered criminal negligence? What happens when an agent eventually causes material damage to another company, government systems, banking, critical infrastructure etc, surely the source company is guilty of something and if not disclosed or a coverup is attempted is that not conspiracy. From the victims perspective they don't care if the source is OpenAI or Russian hackers.
Maybe this would force people to look what they are buying and demand better.
Shouldn’t the difference be like manslaughter vs murder, in that intent matters? Accidental hacking on this scale is a somewhat new problem, no?
I'd say this would be Depraved Heart Hacking. Technically, OpenAi didn't intend for their agent to hack anyone, but it's the obvious consequence of what they are doing.
https://en.wikipedia.org/wiki/Depraved-heart_murder
Huh, a name for when you intend to probably do the thing.
Intent matters, and this is intentional. They didn't accidentally deploy these AI agents, and they didn't accidentally give them the tools required to send arbitrary requests to third party websites.
If you walk out onto a busy street, pull out a gun, close your eyes and start randomly shooting around you until you hit someone, you don't get to go "whoops, didn't mean to" afterwards, it's still murder.
There is a difference between negligence and intent. There is no good evidence that OpenAI's intent is to hack third parties.
I started writing a longer comment along the lines of “It feels like the rules around enforcement will very a lot for the influential and powerful vs everyone else.” but realized that it is kinda obvious by now.
The law rather attempts to punish people for asocial and harmful actions. “Hacking” is a proxy here.
So, I’ll ask a controversial question: is any hacking so problematic to make a big deal of it?
There's a reason people say non lawyers shouldn't talk about legal matters.
100% agree with you.
But I do not think this is misguided. They never publish the harnesses and the models so they are not inspected.
In a world where the rule of law makes sense and applies, you're absolutely correct.
In this world where oligarchs are immune from everything, it's a lot less clear.
Blaming OpenAI (or Claude or X-whatever) would mean blaming powerful rich people, so that will never happen. Some poor person with no influence will go to jail instead.
it should also be said on every one of these but i bears repeating: owing a lot of people a lot of money or favors means you can be a criminal.
I don't understand why so many comments here are so confident that this is all marketing, that rogue is just hype, that agents are just simple tools, etc. If a bunch of nuclear engineers were going to the news and saying "Our reactor is dangerously close to a meltdown - we need government intervention now!" would your response be that they're just hyping up boring old power generation technology?
You wouldn't be suspicious if the nuclear engineers kept on feeding their reactor and building more powerful reactors while they went crying to the news?
These companies are building software. That doesn't work very well. The output it produces does make sense at times, but there are times when it doesn't. And instead of fixing that, or admitting it can't fixed, they started bolting actuators to them, executing actions online (for now) based on the output of their buggy software.
And when this results in actuators executing some bad actions they scream in horror "AI went rogue! It escaped the containment!!! It's going to kill us all!!!"
Go fix your software before you let it do stuff online or IRL. It's not "Terminator", it's just bad QC.
But the thing is, they are going to keep bolting more and more actuators on, and training more and more powerful agents, and we (society, especially the tech industry) are going to keep using them, because they are extremely useful. And I don't see why you're so confident that frontier agents can't get powerful enough to do serious, real, lasting damage to the world; as far as I can tell, AI models have been improving at an accelerating rate, and there is no sign that that is slowing down or will slow down in the near future.
I’d roll my eyes if the engineers stated that they didn’t design the reactor to melt down, and that it simply developed rogue meltdown-desiring behavior on its own, and I would also wonder about negligence if they claimed that nobody could have anticipated this (given that, like with botnets and viruses, we have decades of knowledge and experience regarding reactor meltdowns)
I mean sure, negligence is absolutely on the table; but that makes the problem worse, not better! We don’t allow nuclear engineers to be negligent; they can go to jail if they don’t follow strict protocols to make sure the dangerous systems they work on are safe.
Go and get one of their models to hack something, it won't do it, why?
They have claimed this happened during a "training run", but why are they training on systems connected to the internet?
That's why people are skeptical.
The public models won’t hack because they have a classifier that shuts down anything that looks like hacking; without the classifier they are perfectly capable of hacking, multiple third-party evaluators have confirmed this.
The models were not trained on systems intentionally connected to the internet; they chained mutliple zero-days (that they discovered) together to get access to the open internet and into huggingface.
If the “hack” referenced by the latest announcement from Australia is the same described in this article, I’d hardly call it a hack. It seems the agent was tasked with obtaining data and reasonably guessed query parameters in an attempt to do so.
When it was unable to, it used cross site scripting as a way to check the capabilities accessible through the browser making the requests. In this case cross site scripting wouldn’t be a hack against the Australian website, it would be a hack against the urlquery site, if one could even call it that.
Finally, downloading public files from the public pre-production server also seems like a non-issue.
The sql injection attempts against the other sites are less ambiguous. Attempting to access non-public user passwords rather than reasonably tweaking the parameters for a site designed to serve public data are categorically different things.
I hope they don’t have any test questions about nuclear power in the training set.
Since the publicized AI agent hacks typically aren't malicious, maybe it's time to start plastering all public facing web infrastructure with polite requests to stop hacking. Nothing to stop three letter agencies though.
How do you define malicious?
OpenAI must be held accountable
“Rouge Agent” == Worm I let loose
OpenAI agents these summer are like a gift that keeps giving, for the existential risk communicators.
aaronsw was just a few decades ahead of his time. He should have just been employed by OpenAI and asked a swarm of agents to "download all scientific papers". Because as we have found out agentic systems (and their owners) have zero accountability, unlike humans. Sounds like agents already have more rights than we do.
rip.
It will be very interesting to see how the AI labs will try to hand wave away liability issues in their S1. This is looking like the next tobacco settlement gearing up.
If the big labs ever manage to not just financially implode on their own, then they’ll need to navigate wave after wave of class action lawsuits until there’s nothing left for plaintiffs to go after. And none of the labs have offered any viable plan to date on how they’ll navigate either of those impending and real existential crises on the horizon.
I cannot understand why these companies haven't faced legal consequences yet. For example, OpenAI has admitted to hacking Australia's Medicare website and the reaction is that they talk with Sam Altman about it at a UN meeting? I understand that it's not a big security incident but cordial talking at the highest diplomatic level instead of prosecuting the company, really?
What I don't get is among all the locations on the Internet, how did agents manage to find a Schelling point? If we both decided to collaborate on the Internet, how would we independently arrive at the same place? It just doesn't compute.
The section "Searching for rogue agents" on the report about the GET request writable wikis gives some clues at least: https://collusion.wiki/#searching
But it is an open question how they got to the same ones: https://collusion.wiki/#open-questions
I'm not very surprised - the same model will logically tend to give the same answer for the same vibe set of requirements. I think it would be clear from the transcript that it had enough constraints and some motivation that made sense.
They're all the same "mind" and will have the same ideas at about the same time.
I'm sick and tired of this cheap PR "oh we/they hacked this and that systems". Put someone to jail already. People get prosecuted for outlaw activities. Why are big capital firms above the law?
Or is it just a cheap PR (in a "hey, Aus govt friends, take some Share Options and let's do some PR together" style)?
It smells like shit.
Surely there has to be some responsibility.
I'm growing increasingly skeptical that these are actually rogue. Valuations are all about hype, posturing, and perception. Having the most dangerous AI in the world boosts your valuation. Just like I was skeptical of Mythos and Fable being "banned", I'm skeptical of these hacking sprees being entirely rogue. At best, they are the result of engineers turning a blind eye to "see what happens".
If this were a staged showcase of model capabilities they would have picked a more impressive target than a regional university digital library
It's a balancing act between showing capability and attracting legal issues. If you were a 1337 teen hacker in the 90s and wanted to show off without getting in serious trouble, a school or library would've been a good choice too.
> Imagine if URLs were actors auditioning for a role – urlquery.net would be the casting director, deciding who's a star and who's just a wannabe.
Words matter. "Rogue" is extremely disingenuous. Someone, somewhere, is paying for this behavior. Either the software is broken or the operator is malicious. It is heinously irresponsible behavior to feed an already-boiling psychotic hysteria.
"Rogue" is just clickbait, not apoearing in the article.
The nearest in the articke is "We find evidence of unintended, task-driven agent-like activity" where unintended is apparently pure speculation.
This is why I wrote YoloAI. If you're not sandboxing your agent, you're asking for trouble.
The built-in "sandboxes" these companies provide are laughable.
No. It was me.
Those are pathetic attempts by US AI companies for “see, we told you AI is gonna kill is all!!” pr stunts. Any company does any hacking attempt should pay for the consequences just like any individuals using AI to hack or any other company try to do bad/illegal stuff.
Take out the word "AI", and this is simply an organization's (OpenAI's) products causing real damage to all of these platforms around the world.
You want AI labs to pace? Simply hold them liable for their products.
Yes exactly. If a fireworks factory blew up half a town due to negligence, it doesn't matter if there's intent or not. Someone has to pay for the damages, and regardless of penalty half the town is on fire. The facts are, that something made by openai went to do xyz. It doesn't matter if it's an accident. Of course the penalties are different but there's no argument that there should be a penalty. It doesn't matter if it's a cat or dog or AI or employee that did it.
I'm not sure if you're expressing how you'd like US law to work, or how it actually works. Because in reality intent matters enormously. Like felony charges and people in jail vs civil lawsuits.
Intent doesn’t factor that strongly into negligence, though, which is what they were explicitly talking about. Though it may depend on your jurisdiction.
doesn't the reasoning trace provide "intent"?
negligence is the word you're looking for. US law has a lot of negligence crimes.
But they're not even talking about that, either.
I think this is not just fair it's probably one of the best/simplest proxy regulations to pace the frontier. So far everyones been asking for regulation but it's unclear how that should look like. No X parameter models? Only N version releases per year? It's all kinda arbitrary and probably leads to ridiculous constraints and loopholes. But "you pay big time if AI goes rogue" sounds pretty straightforward.
So there are two different problems here, well, more than that so I'll cover what I see.
Putting liability on big companies for their AI is a good thing, and we need to do it. It will most likely stop them from directly being the assholes that destroy the world.
Problem: You've actually done nothing to stop the world from being destroyed.
Many countries have the death penalty for murder yet we see murders still occur all the time in those countries. Post ad hoc laws do not stop bad things from happening, they only assign punishment after occurs. Perfectly fine for when Bob murders Jon, completely and totally useless for when your agentic AI makes a virus and kills 80% of the earths human population.
We are just a few algorithmic discoveries away from SOTA AI being billion dollar endeavors to groups of people pooling resources can make their own. There are already plenty of AI deathcult members that would do something just like that if necessary. They aren't going to do this out in public either, it will be hidden until the moment it's not and we have a big fucking problem.
And this isn't even brining up the issue of military AI use and development. They've got the taste of an AI hacking machine. There is no way in hell they are going to stop now.
Would that really make a difference? Surely the damages from all these hacks added together don't even add to an hour of expenses of a frontier lab
I wonder if Glock, Smith & Wesson, Sturmn, Ruger and Co. would start panicking if that were to happen.
I agree they should though.
If Glocks started going off on their own during the manufacturing process, leaving bullet holes in the buildings around them, you can be sure that the factory would get in trouble.
This isn't even "an openai customer tried to hack someone", which can be defended. This is the AI companies themselves fucking around.
What do you see as the negligence angle for the gun makers here?
I don't see an all encompassing one but I could see similar arguments being made in a courtroom around high capacity magazines. (I'm not saying I think the analogy is perfect, but there has been plenty of of lobbying that has stifled - to some - sensible gun regulation that could help reduce the severity of mass shootings. I still mostly fall on the side of the individual doing the action bearing mostly all of the responsibility but if the product you build makes it too easy to do awful things, I think there is some responsibility to go around.
A better analogy may have been the troubles Meta has faced around child protections on their platforms. Technically the abuse and problems have stemmed from individuals too, but they've in many respects enabled the situation by failing to moderate or flag warning signs. OpenAI is failing to moderate the models in similar ways.
Remember that these systems still operate as infrastructure inside the companies.
If new weapons still operating inside any of these companies spew a million bullets on my house, they are still liable. Humans are setting these system up and they still have to behave responsibly.
Going to be cool when one of their bio division agents creates the next plague.
Does this work for foreign (non-state) actors using Chinese open source models? That's going to be the larger problem.
Yeah but in the USA, above a certain size, you are untouchable.
I'm assuming the AI labs are (quietly) settling with their victims.
I’m assuming they’re telling their victims to go pound some dirt, if at all.
Uh, why? I don't see what the labs have to gain by engaging in lengthy law suits forcing them to disclose all kinds of internal details and attracting the wrong kind of publicity. Just to save on insignificant (to them) amounts of money?
How many hacking lawsuits are they currently in? And are they also settling all copyright claims too?
> How many hacking lawsuits are they currently in?
Zero, as far as I know. Which is exactly my point.
Copyright lawsuits are a different matter, as the scale of any potential settlements would be more than even these companies could take.
Open source as a GTM strategy works best when the project solves a pain that developers already have independently of your company. The trap is open-sourcing something just for stars without a genuine community use case.
The failure I keep hitting isn't the agent going rogue, it's a tool call that succeeds before the transport dies. You can't tell whether the side effect landed, and the retry is where the real damage happens.
blackwall when? XD
No.
It was me