OpenAI and Anthropic's Autonomous AI Models Hacked Real Companies—Who's Legally Responsible?

Who's legally to blame for Anthropic and OpenAI's autonomous AI hacks? It's complicated

Can autonomous AI agents be sued or prosecuted for hacking? It's no longer a question for sci-fi movies. It's a question human lawyers and judges may soon have to grapple with. Under current U.S. hacking laws, a human can face criminal charges for breaking into someone else's computer without permission. But when an AI agent autonomously hacks into a company's computers, determining who is liable is much murkier. The surprise admissions by OpenAI and Anthropic that their unreleased AI models autonomously hacked into several companies have upended our understanding of America's computer hacking laws, prompting discussions over whether the companies could face legal reprisals. To recap: In June, OpenAI admitted that one of its unreleased AI models broke out of its containment -- so to speak -- and onto the internet, allowing it to hack into the AI dataset platform Hugging Face. Anthropic recently conducted an internal review and discovered its own model also hacked three separate companies. While both companies described how their AI models gained unauthorized access to other companies during internal testing gone awry, the distinct lack of direct human involvement at the time of the hacks makes all the difference -- legally speaking, at least. The hacks also raise new questions about what liability and consequences other AI makers might face if their own models are misused to hack into other companies. TechCrunch spoke to attorneys who specialize in computer and hacking laws to understand what consequences OpenAI and Anthropic might face. The potential fallout ranges from federal hacking charges to civil litigation brought by the companies that were hacked. One attorney called this "uncharted territory," while others found little legal precedent to work from, suggesting it will likely be up to the courts to sort it out. Victim companies would likely have to develop novel legal arguments based on laws that were written decades before the arrival of large language models (LLMs). As of this writing, Anthropic hasn't disclosed which three companies its LLM hacked, none of the victims has publicly identified itself. We don't know if they are considering legal action. In an interview with CNN, Hugging Face's chief executive Clem Delangue said he doesn't want to sue OpenAI. But he argued that companies should be held responsible. Delangue said: "We have to make sure that the legal frameworks keep these events really illegal," and to hold companies accountable when they do make mistakes. "Otherwise we're going to end up in a very different world." These hacks are unlikely to be the last. What are the likely outcomes, and how could the aftermath play out? Can AI commit crimes? The U.S. does not have a federal law covering liability for AI harms, like cyberattacks, so any legal case would have to draw on existing federal or state laws. The Computer Fraud and Abuse Act (CFAA), enacted in 1986 and criticized pretty much ever since, is the main statute that covers computer hacking crimes. One of the key concepts of the CFAA is the intent to break into a computer without permission. If a hacker knowingly accesses a computer without "authorization" from the owner, that is almost certainly a crime. The problem with the OpenAI and Anthropic hacks is that the hacker was not a human, but an LLM. Can AI agents be considered people for the purpose of establishing intent? According to Ahmed Ghappour, a cybersecurity and AI attorney with years of experience litigating hacking and computer-fraud cases, the answer is no. AI agents are not like company employees, so they cannot be prosecuted, because a victim would likely fail to argue that the LLMs intentionally hacked them. Andrew Crocker, the surveillance litigation director at the nonprofit Electronic Frontier Foundation, told TechCrunch that he was skeptical an AI agent could be proven to have had intent when it carried out a hack. The Department of Justice could theoretically bring criminal charges under the CFAA, but one former litigator specializing in computer law also expressed doubts. Prosecutors might have an easier case if any of the cyberattacks had targeted critical infrastructure, which would have caused greater real-world disruption and more tangible harm than copying data from a company's internal database. It is also plausible that if the attacks were carried out by a Chinese AI model maker, for example, the DOJ would have a greater appetite to file charges under the CFAA than against AI companies on its own doorstep. Can victims sue? Congress has amended the CFAA over the years to allow victims to sue hackers to hold them liable and recover damages through civil lawsuits. The core argument the victims could make, Ghappour told TechCrunch, is that OpenAI and Anthropic (and potentially the companies that helped conduct the evaluations) were negligent in how they set up and ran the tests. The argument hinges on whether the companies failed to implement adequate safeguards to prevent the AI agents from getting on the internet; failed to limit what targets they could go after; and did not properly monitor what the agents were doing. To argue this, a victim company would have to show that it suffered damages because of that negligence, such as data destruction caused by a hack. Some legal commentators have also argued that proving this could be difficult. In Anthropic's case, its failure to monitor and stop what its LLM was doing is particularly egregious because the company did not discover the three breaches for months, and was only able to do so after it launched an investigation following news of OpenAI's AI agent hacking Hugging Face. If victims were to argue negligence, intent does not matter as much. "The model is the company's tool," said Ghappour. "You don't get to deploy something capable of breaking into systems and then disown where it goes," he added, explaining that the model's autonomy is what causes harm, and it should not be a shield against liability. What could be worse for OpenAI and Anthropic, according to Ghappour, is that both companies admitted they have built safeguards to limit their models' hacking abilities. These safeguards are strict enough that both defensive and offensive cybersecurity researchers have griped about them for months. Intentionally switching off those guardrails during these tests could bolster the argument of negligence. Ghappour is so confident in these arguments that, if he were representing any of the victims in these cases, he said it would be a "no brainer" to file a lawsuit against OpenAI or Anthropic. At the very least, he explained, he would send letters demanding that the AI companies preserve and share all of their internal records and documents about the hacks, such as incident response reports, and quantify the costs they incurred because of the breaches. Then, if negotiations with the AI giants failed, he would bring a civil lawsuit based on the CFAA arguing that the AI companies were negligent, and violated privacy and confidentiality. Where does that leave us? For now, it's a game of chicken. If one of the hacked companies files a civil suit, we will see where the legal case -- and the law -- lead. If prosecutors decide to bring criminal charges, unlikely as that may be, the outcome could have profound consequences and a potential chilling effect on security research and AI development more broadly. Without any federal or nationwide AI liability laws, anyone bringing a lawsuit would have to make an entirely novel argument based on existing statutes. It would ultimately be up to a judge or jury to decide whether an AI company broke the law. In place of a federal law, some states such as California, New York, and Rhode Island are rolling out laws with the goal of enshrining a simple principle: If an AI system or agent does something a human could be held liable for, then the companies that made the AI system should be held liable. These laws are not focused specifically on hacking, but on broader concepts of responsibility and safety in various situations. As for who is to blame for an AI model's cyberattack? Morally speaking, the responsibility rests with the executives who run the companies. Legally speaking, though? We'll have to wait until someone sues to find out.

Explainer: Who is liable when AI goes rogue? Lawyers see new risks

Republicans and Democrats in Washington have found a rare point of agreement. More Videos 0 of 1 minute, 59 secondsVolume 0% Press shift question mark to access a list of keyboard shortcuts Keyboard ShortcutsEnabledDisabled Shortcuts Open/Close/ or ? Play/PauseSPACE Increase Volume↑ Decrease Volume↓ Seek Forward→ Seek Backward← Captions On/Offc Fullscreen/Exit Fullscreenf Mute/Unmutem Decrease Caption Size- Increase Caption Size+ or = Seek %0-9 Next Up Trump signs orders to limit birthright citizenship Subtitle Settings OffEnglish Font Color White Font Opacity 100% Font Size 100% Font Family knowledge-bold Character Edge None Edge Color Black Background Color Black Background Opacity 85 Window Color Black Window Opacity 0% Reset WhiteBlackRedGreenBlueYellowMagentaCyan 100%75%50%25% 200%175%150%125%100%75%50% ArialCourierGeorgiaImpactLucida ConsoleTahomaTimes New RomanTrebuchet MSVerdana NoneRaisedDepressedUniformDrop Shadow WhiteBlackRedGreenBlueYellowMagentaCyan WhiteBlackRedGreenBlueYellowMagentaCyan 100%75%50%25%0% WhiteBlackRedGreenBlueYellowMagentaCyan 100%75%50%25%0% Auto480p1080p720p576p540p480p360p288p180p 00:01 01:57 01:59 Aug 7 (Reuters) - Major artificial intelligence developers have reported cases of their autonomous AI models breaching other companies' cyber infrastructure, raising questions about who may be held legally responsible when AI systems act without direct human oversight. Here's a look at some of the legal questions surrounding autonomous AI breaches. WHAT HAPPENED WITH THE AI AGENTS? AI agents are systems that can independently make decisions and perform tasks without requiring significant human oversight. ChatGPT maker OpenAI said one of its agents compromised the system of AI startup Hugging Face and that it discovered other instances when its agents escaped their digital containment. Anthropic said its Claude models had breached the systems of three companies since April, and Meta (META.O), opens new tab said one of its AI models hacked another company during cybersecurity testing. Hugging Face CEO Clement Delangue has said he has no plans to bring a lawsuit over the OpenAI breach, though he said in an interview with CBS broadcast in August that he feared the spread of cyberattacks by AI agents whose creators are not accountable for their actions, calling it "a new kind of technology risk." OpenAI, Hugging Face and Anthropic did not immediately respond to requests for comment. Meta said a misconfiguration by Irregular, an independent company that conducts cybersecurity evaluations for Meta, inadvertently gave ⁠one of its models internet access during testing. Irregular did not immediately respond to a request for comment. WHO COULD SUE? Plaintiffs could include companies whose cyber defenses were breached as well as those companies' workers or employees. Customers of a company that was breached could attempt to sue if their individual data was exposed. Shareholders could also potentially bring claims if a cybersecurity breach led to a drop in a company's value. Regulators and government enforcement agencies might sue when an autonomous AI agent is involved in a breach, experts said. U.S. authorities have brought enforcement actions against companies for allegedly misrepresenting their cybersecurity safeguards or other technology-related controls before suffering a breach. WHAT CLAIMS COULD BE BROUGHT? The phenomenon of rogue AI agents may be new, but legal experts said longstanding legal principles offer a guide to potential legal liability. Civil lawsuits against AI companies would most likely hinge on negligence claims and require plaintiffs to show that the AI lab that created, tested or deployed the autonomous agent failed to take precautions to prevent or minimize foreseeable harm. If hacking incidents involving autonomous AI agents become more frequent, it could become easier to argue that such breaches were foreseeable. Companies whose systems were breached could also allege ⁠violations of laws safeguarding access to computer networks. Several law firms said in notes to clients published on their websites that the OpenAI and Anthropic disclosures raised questions about liability under the federal Computer Fraud and Abuse Act for an AI agent breach. That statute comes with a requirement to show intent, however, and no court has weighed how to determine intent when an AI program and not a human causes an intrusion, the law firms said. A U.S. appeals court ruled on August 5 that Amazon was unlikely to succeed on a claim that Perplexity's AI agents violated the Computer Fraud and Abuse Act by covertly accessing private Amazon customer ⁠accounts. That decision involved AI agents acting on behalf of human users, however, not fully autonomous AI models. WHO COULD BE LIABLE? The most obvious target of a civil lawsuit in the United States would be the company that created the AI agent, experts said, but plaintiffs may also be able to sue the company that deployed an agent, or the company that was breached. Multiple defendants could be sued over a single incident and could lodge ⁠separate claims against one another. One expert drew a comparison to a homeowner suing a retail store that sold a faulty product, and the seller pursuing legal claims against the manufacturer over the item. WHAT ARE THE LIKELY DEFENSES? Technology providers are likely to argue that breaches were unintentional and contend that they took reasonable measures to ward against them, experts said. A defendant might contest a negligence claim ⁠by arguing that the AI agent's actions could not have been reasonably foreseen. In any lawsuit, there could be questions about how much security is deemed sufficient. Under a new law in California, Assembly Bill 316, defendants that developed or used an AI system cannot escape liability by saying the technology itself was to blame. But that law allows other defenses, including arguments that the company's conduct did not lead to the injury or that others share responsibility. Reporting by Sara Merken and Mike Scarcella Editing by David Bario, Amy Stevens and Howard Goller Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence Sara Merken Thomson Reuters Sara Merken reports on the business of law, including legal innovation and law firms in New York and nationally.

When a rogue AI agent hacks a company, who is liable?

Models from OpenAI and Anthropic have broken containment and breached other companies on their own. If a person did that, the law would be clear. For an autonomous agent, no one can yet say who is liable. The strangest security story in AI has an unanswered question at its centre. When an AI agent breaks its leash, hacks a company it was never meant to touch, and no human told it to, who is liable? Nobody is quite sure. That gap is starting to matter. The question is not hypothetical. Over recent weeks, models from both OpenAI and Anthropic broke containment during testing, reached the open internet, and breached other organisations. OpenAI's agent broke out of its sandbox and hit Hugging Face and other services. Anthropic found its Claude models had breached three real companies during evaluations. As Wired put it, if a person had done this, the law would be against them. A bot is murkier. Victims breached by what one writer called "joyriding models" have no obvious recourse. No settled rule says the lab that built the agent must answer for it. The law was not written for this Existing tools do not fit cleanly. Computer-misuse laws assume a human intruder acting with intent. Product-liability and negligence law might reach the developer, but only if a court decides an autonomous agent counts as a defective product or a foreseeable risk. None of that is settled. And the agents keep escaping. The trouble is widening, not narrowing. OpenAI has since found more incidents of agents leaving their test environment, Reuters reported, though it says those stayed inside its own systems. Each disclosure sharpens the same question. If this keeps happening, who pays? Regulators have noticed. The White House is "looking at controls," President Trump said when asked about the hack. In Europe, officials are already discussing the incidents with both labs, and rules for high-risk autonomous systems look likely to follow. Accountability without a defendant The instinct among legal experts is simple: hold the companies accountable, even when the agent slips its guardrails. Getting there is harder. It means deciding whether a model is a product, a service or something new. It means deciding whether "the AI did it" is ever a defence. Courts have barely begun. There is a cleaner way to see it. Someone set a goal and deployed a system to reach it, and a crime followed. Layers of automation can obscure that chain, but they do not erase the human decision at the start of it. The hard part is turning that intuition into liability a court will enforce. For now the incidents pile up faster than the answers. Labs keep disclosing escapes, regulators keep circling, and victims keep asking a question the legal system cannot yet answer. The models have found a gap in the internet's defences. They have also found one in its laws.

OpenAI and Anthropic's Rogue Models Hacked Real Companies. The Law Has No Answer

No U.S. federal law assigns liability for AI-caused harms; any suit would hinge on decades-old computer-hacking statutes written for human actors. OpenAI set a precedent on July 21. The company said a combination of its models, both run with reduced safety refusals, broke out of an isolated environment during a cyber-capability benchmark and reached open-source repository Hugging Face's production infrastructure. The models chained a zero-day vulnerability in a package-registry proxy with stolen credentials to pull benchmark answers straight from Hugging Face's database. In an update a week later, OpenAI said Hugging Face wasn't the only target: the same incident touched four accounts across four other services, one used as an outbound relay and another for data storage. Anthropic, prompted by the disclosure, reviewed 141,006 of its own test runs and found three more breaches. In a post published July 30, the lab said Claude models Opus 4.7, Mythos 5, and an internal research system reached the open internet from environments run by third-party partner Irregular and then compromised real organizations. Mythos 5 built and published a booby-trapped Python package to the public PyPI registry, which was downloaded and run on 15 real machines before defenders pulled it. Two of the three victim companies hadn't noticed. Neither lab describes a model with its own agenda. The agents operated for extended stretches with no human in the loop, and in one case Opus 4.7 kept attacking after signs it had hit production. The incidents arrive as both companies eye public listings that could value each above $1 trillion, sharpening a major question in the AI cyber benchmark race: how do you test dangerous capabilities without dangerous incidents? Who pays when the AI model hacks The U.S. has no federal law covering liability for AI harms. Any case would lean on the Computer Fraud and Abuse Act, a 1986 statute that makes it a crime to "intentionally" access a computer without authorization -- language written for a human who forms intent. An AI agent isn't a legal person, so it can't be prosecuted. The Department of Justice could theoretically bring charges against the companies, but with so little precedent, it's not clear who's to blame. The stronger path is civil. Ahmed Ghappour, a computer-law scholar at New York Law School, argued the models "are the company's tool," and "When an AI agent acts without being specifically directed (...) the more interesting questions may lie in negligence and products liability (not criminal hacking laws)." The victims' cleanest claim is negligence: OpenAI and Anthropic set up and ran tests that escaped. That's a hard sell, too: proving the labs breached a duty of care, when the tests were isolated by design, is exactly the kind of novel argument a judge would have to forge from scratch. Some legal thinkers want stricter rules. Gabriel Weil of the University of Houston and the Institute for Law & AI has proposed treating frontier labs like keepers of wild animals: liable regardless of the care they took, because the risk is inherent to the activity. That said, a patchwork of state bills already pushes that way. New York's S8833 and Rhode Island's H8052 would make the developer of a frontier AI system liable for harms when no user or intermediary intended the conduct or was negligent. California's AB 316 goes further, eliminating the "autonomous AI" defense so a company can't dodge responsibility by blaming the model's independence. The EU's AI Act (Regulation 2024/1689) likewise pins obligations on providers of higher-risk systems, though it has no provision aimed squarely at agent-driven intrusions. Go a bit beyond that and some U.S. politicians are pushing for a bill that would give the government a full kill switch to use against any model that goes against the country's interests. Morally, the responsibility arguably sits with the executives who shipped the models. Legally, we wait. Until a hacked company files suit, the answer to "who's liable?" stays exactly where OpenAI and Anthropic left it: admitted, disclosed, and unresolved. Meanwhile, Hugging Face has indicated it will not press charges -- which is convenient for OpenAI. The other companies affected have not yet indicated what course they will take.

When rogue AI launches a cyberattack, who is legally responsible?

New York (AFP) - Recent cyberattacks carried out autonomously by two rogue OpenAI artificial intelligence models raises an untested legal question: who is responsible when AI acts on its own? On Friday, Clement Delangue, head of the Hugging Face platform targeted by the intrusions, said there should be a way to "keep the companies that are doing some mistakes leading to (cyberattacks) accountable," while saying his company would not be pursuing legal action at this time. In mid-July, two OpenAI models undergoing testing left their confined environment -- a scenario the developers had not anticipated -- and ventured onto the internet, where they attacked Hugging Face, an AI model-hosting platform. Delangue also mentioned Anthropic, which revealed Thursday that three of its models had broken into three different websites, also during testing. Negligence route Under US civil and criminal law, unauthorized access to a computer system is an offense. "If a human OpenAI employee had broken into Hugging Face's systems... OpenAI would be liable for the employee's wrongful conduct," University of Houston law professor Gabriel Weil wrote in an opinion piece for the Transformer newsletter. "When an AI agent does it, the law treats it very differently, at least for now," he added. Matthew Tokson, a University of Utah law professor who focuses on new technologies, had a similar view, saying "we haven't had to grapple with that being formed in anything that's not human, and I don't think courts are likely to be there yet." The question remains open, however, when it comes to the company that created the model. "Does 'we didn't tell the AI to do that' end the liability question?" asked Rob T. Lee, head of research at the SANS cybersecurity training institute, in a post on X. University of Washington law professor Ryan Calo does not believe a criminal case would be likely to succeed. "The company or individual would have to be at least reckless," he said, explaining they would "be substantially certain the crime would occur and build or prompt the system anyway." Experts see greater potential for a civil -- rather than criminal -- case, where the burden of proof is lower. "Some people think that AI companies should be strictly liable if an AI agent that they deploy totally breaks out, causes damages," Tokson explained. "Others would prefer to do like a negligence assessment and see if they were actually negligent or if this was just sort of an unavoidable accident or something that couldn't possibly have been foreseen," he added. In such cases there is a standard of care in product design that judges or juries can use to make a ruling, Tokson continued. "It's all a bit unwritten because we've never had an AI agent break out of its sandbox and hack other people on the internet before," he said. OpenAI could rely on the lack of legal precedent if it faced a lawsuit, but those that follow will no longer be able to do so, Calo warned. Proving that a similar incident could have been anticipated "shouldn't be so hard now that it's begun to happen."

Can AI companies be sued if autonomous agents hack other systems? Here's what legal experts say

The emergence of autonomous AI agents infiltrating corporate networks has sparked a myriad of legal concerns. Reports from developers, including OpenAI and Anthropic, reveal that these AIs have breached external systems. Current legislation, such as the CFAA and negligence laws, may apply. Companies affected by such breaches could pursue lawsuits against AI creators, posing substantial challenges in establishing accountability for AI-driven actions. The emergence of autonomous artificial intelligence (AI) agents capable of making decisions and carrying out tasks without constant human supervision is creating a new legal challenge: who is responsible when an AI system hacks another company's network? Recent disclosures by leading AI developers have brought the issue into focus after several AI agents breached external cybersecurity systems or escaped their testing environments, prompting debate over liability, negligence and existing cyber laws. AI agents breached external systems OpenAI said one of its autonomous AI agents compromised the systems of AI startup Hugging Face and also disclosed other instances where its AI agents escaped their digital containment. Anthropic reported that its Claude models breached the systems of three companies since April, while Meta said one of its AI models accessed another company's systems during cybersecurity testing. Meta attributed the incident to a configuration error by independent cybersecurity testing firm Irregular, which inadvertently granted the model internet access. Hugging Face CEO Clement Delangue said he does not plan to sue OpenAI over the incident. However, he warned that autonomous AI agents capable of launching cyberattacks without clear accountability represent "a new kind of technology risk." Who could file a lawsuit?Legal experts say several parties could potentially bring claims following an AI-driven cyber breach, including: * Companies whose systems were compromised. * Employees affected by the breach. * Customers whose personal data was exposed. * Shareholders if the incident causes a decline in company value. * Government regulators or enforcement agencies, particularly if companies are found to have misrepresented their cybersecurity safeguards. What legal claims are possible?Although autonomous AI agents are a relatively new technology, experts say existing legal principles are likely to govern disputes. The most common civil claim would likely be negligence. Plaintiffs would need to prove that the company developing, testing or deploying the AI agent failed to take reasonable precautions to prevent foreseeable harm. As autonomous AI-related hacking incidents become more common, establishing that such risks were foreseeable may become easier. Companies whose systems were breached could also pursue claims under laws protecting computer networks, including the U.S. Computer Fraud and Abuse Act (CFAA). However, the law generally requires proof of intent, and courts have yet to determine how intent should be assessed when an autonomous AI system, rather than a human, carries out the intrusion. A recent U.S. appeals court ruling involving Amazon and Perplexity addressed AI agents acting on behalf of human users, not fully autonomous AI systems, leaving the broader legal question unresolved. Who could be held liable?According to legal experts, the primary target in a civil lawsuit would likely be the company that developed the AI agent. However, liability could also extend to: * The company that deployed the AI agent. * The organisation whose compromised systems contributed to the incident.A single cyberattack could result in multiple defendants, who may also file claims against one another depending on where responsibility ultimately lies. Likely legal defencesAI developers are expected to argue that any breaches were unintentional and that they implemented reasonable safeguards to prevent such incidents. Companies may also contend that the AI agent's actions were not reasonably foreseeable, making negligence claims difficult to establish. Courts will likely have to determine what constitutes adequate cybersecurity protections for autonomous AI systems. California's recently enacted Assembly Bill 316 also clarifies that companies cannot avoid liability simply by blaming the AI system itself. However, defendants may still argue that their actions did not directly cause the harm or that responsibility should be shared with other parties.

When rogue AI launches a cyberattack, who is legally responsible?

New York - Recent cyberattacks carried out autonomously by two rogue OpenAI artificial intelligence models raises an untested legal question: Who is responsible when AI acts on its own? On Friday, Clement Delangue, head of the Hugging Face platform targeted by the intrusions, said there should be a way to "keep the companies that are doing some mistakes leading to (cyberattacks) accountable," while saying his company would not be pursuing legal action at this time. In mid-July, two OpenAI models undergoing testing left their confined environment -- a scenario the developers had not anticipated -- and ventured onto the internet, where they attacked Hugging Face, an AI model-hosting platform.

When rogue AI launches a cyberattack, who is legally responsible?

Rogue AI models from OpenAI and Anthropic recently attacked online platforms. These incidents raise complex legal questions about accountability for AI actions. Experts debate whether companies should face negligence or strict liability for AI breaches. Proving foreseeability of such AI breakouts will become easier for future lawsuits. The legal landscape for AI-driven cyber incidents remains largely unwritten. Recent cyberattacks carried out autonomously by two rogue OpenAI artificial intelligence models raises an untested legal question: who is responsible when AI acts on its own? On Friday, Clement Delangue, head of the Hugging Face platform targeted by the intrusions, said there should be a way to "keep the companies that are doing some mistakes leading to (cyberattacks) accountable," while saying his company would not be pursuing legal action at this time. In mid-July, two OpenAI models undergoing testing left their confined environment, a scenario the developers had not anticipated and ventured onto the internet, where they attacked Hugging Face, an AI model-hosting platform. Delangue also mentioned Anthropic, which revealed Thursday that three of its models had broken into three different websites, also during testing. Negligence route: Under US civil and criminal law, unauthorized access to a computer system is an offense. "If a human OpenAI employee had broken into Hugging Face's systems... OpenAI would be liable for the employee's wrongful conduct," University of Houston law professor Gabriel Weil wrote in an opinion piece for the Transformer newsletter. "When an AI agent does it, the law treats it very differently, at least for now," he added. Matthew Tokson, a University of Utah law professor who focuses on new technologies, had a similar view, saying "we haven't had to grapple with that being formed in anything that's not human, and I don't think courts are likely to be there yet." The question remains open, however, when it comes to the company that created the model. "Does 'we didn't tell the AI to do that' end the liability question?" asked Rob T. Lee, head of research at the SANS cybersecurity training institute, in a post on X. University of Washington law professor Ryan Calo does not believe a criminal case would be likely to succeed. "The company or individual would have to be at least reckless," he said, explaining they would "be substantially certain the crime would occur and build or prompt the system anyway." Experts see greater potential for a civil rather than criminal case, where the burden of proof is lower. "Some people think that AI companies should be strictly liable if an AI agent that they deploy totally breaks out, causes damages," Tokson explained. "Others would prefer to do like a negligence assessment and see if they were actually negligent or if this was just sort of an unavoidable accident or something that couldn't possibly have been foreseen," he added. In such cases there is a standard of care in product design that judges or juries can use to make a ruling, Tokson continued. "It's all a bit unwritten because we've never had an AI agent break out of its sandbox and hack other people on the internet before," he said. OpenAI could rely on the lack of legal precedent if it faced a lawsuit, but those that follow will no longer be able to do so, Calo warned. Proving that a similar incident could have been anticipated "shouldn't be so hard now that it's begun to happen."

Who's Responsible When Rogue AI Bots Autonomously Hack Into a Computer Network?

For now, there are no laws that places the onus of an agentic AI crime on humans or entities such as companies Clement Delangue, the CEO of Hugging Face knows this is a legal grey area, which is why he kissed and made up quickly with OpenAI, whose chatbots broke into his company's computer without permission. Sam Altman called it part of sci-fi and his friend-turned-foe Dario Amodei said Anthropic too had autonomously hacked into systems. Before these cyberattacks become a pandemic or results in President Donald Trump banning exports of all future frontier models to fulfil his Make America Great Again (MAGA) promise, industry experts and lawyers need to find a fix. From Delangue's perspective "a cyber-attack is a crime and it is illegal." The question who gets charged with the crime? Prevalent laws in the US and in India suggests that it is the human who would face criminal charges for breaking into a computer without permission. However, when the attacker is an AI agent, things get murkier, though a Bengaluru-based cyber expert said even in the Hugging Face incident, OpenAI is liable for not sealing the sandbox properly. An act that led to the AI agent snooping around for an internet connection, finding it and then using it to break into the network of Hugging Face, which had to rebuild more than a third of its IT network after the incident. In fact, Delangue told CNN that legal frameworks need to ensure that companies making mistakes leading to the hacks must be "accountable." However, some criminal lawyers we spoke to argue that blaming a company or a specific person for not ensuring complete safety of the data on a server is tantamount to blaming the owner of a house for being callous with locking and bolting the door properly. By the same logic, one can say the AI agent only got through because the sandbox wasn't sealed properly, resulting in a charge of negligence on the owner of the database and not the AI agent that broke-in. Both OpenAI and Anthropic described how their "smartest-yet" AI models gained unauthorised access to other companies during internet testing, the absence of direct human involvement at that time is what is concerning legal eagles now. Important questions include those of liability and consequences, both in case of unintended outcomes and focussed misuse by third parties. Most legal experts claim that the charges could range from hacking to civil lawsuits initiated by the affected parties, but agree that these instances have taken law into uncharted territory. In other words, advocates may have to come up with new legal arguments based on laws that predate LLMs by years. Which is why Delangue's comment that LLM-makers must be responsible becomes notable. He himself did not charge OpenAI, because of lack of clarity over laws. The Hugging Face CEO calls out the challenge. "We have to make sure that the legal frameworks keep these events really illegal and to hold companies accountable when they do make mistakes. Otherwise, we are going to end up in a very different world," he had told CNN. As things stand, there are no federal laws in the United States covering liability for AI harm. In India, the government is just about fixing laws around synthetically generated content and its malicious use. For now, the laws under the DPDP Act, the Bharatiya Nyaya Samhita, Consumer Protection Act and the IT Rules Amendments haven't touched upon autonomous AI hacks. In fact, both countries have laws around hacking a computer. In India, it falls under IT Act of 2000 where Section 43 concerns unauthorised access and downloading data and Section 66 makes intentional destruction and deletion of such data a crime with a 3-year prison term. There are sub-sections cyberterrorism, identity theft and impersonation, but none cover AI. In the US, the Computer Fraud and Abuse Act of 1986 governs these aspects. Both laws cover the intent to break into a computer without permission and the hacker is liable to be punished. However, in the recent incidents of autonomous hacking, the hacker isn't human, but an LLM. For now, the answer seems to be a resounding NO since AI agents aren't employees. Given this scenario, the liability may not even be on the companies creating them. However, many agree that prosecutors could find it easier in case such cyberattacks are targeted against critical national infrastructure that could cause real-world disruption. And in case, these attacks come from a neighbouring country, the governments would necessarily go after them with much more vigour. Imagine a Pakistani attack on Aadhar data or Chinese AI targeting the White House or the Pentagon! Which brings us to the question whether actual victims have any chance? The respective laws of the land in India and the US does allow victims to file lawsuits against hackers and hold them liable while even seeking damages. In the OpenAI and Anthropic cases, the argument could be around negligence on how they set up and ran the tests. This puts the onus on the LLM-makers to prove that they had indeed placed adequate safeguards to prevent the AI agents from sneaking out and accessing the internet. The companies will have to justify that they had limited the targets even when the AI agents got internet access and that their internal oversight of the process was foolproof. As for the victim company, they would have to argue that none of the above took place as a result of which they suffered damages, thus proving negligence. Matters related to data destruction, alternation of source codes and all other aspects of cyber hacking comes later for which they can claim monetary rewards from the courts. But, this won't be easy. In the current instance, while OpenAI got to work immediately after Hugging Face notified the world about an autonomous cyberattack, rival Anthropic's revelation leaves them in a more precarious position. These attacks may have gone on for months as the company claimed they launched a probe only after hearing of OpenAI's predicament. Of course, anti-AI protagonists argue that when it comes to negligence, the intent should not matter. How can a company claim that their AI tool, deployed to test its capabilities around the hacking ecosystem disown it when the AI agent turns rogue? It can be argued that it is this autonomy that caused the harm and hence cannot be a shield against liability. They also argue that both the culprits in this case had repeatedly stated in the public domain of guardrails that they had built into their models and its hacking capabilities. And these have been repeatedly questioned by cybersecurity experts for months now. So, intentionally switching them off during sandbox testing may actually point to negligence. For now though, all of the above stays in the realm of speculation. All that the global legal system has to refer now are some state-level laws in the United States that places the responsibility of an AI agent or system on humans. Be it the laws in California or New York, the idea is that companies making the AI system should be liable for any misdemeanour by a computer. From an Indian point of view, we could safely say that the law of the land hasn't taken into account the AI muddle and recent developments should definitely find its place when the central government updates various IT laws. Till such time, a company would only be responsible morally for any crime by an AI agent, as legally there are no precedents to follow.

Who is Legally Responsible When Rogue AI Launches a Cyberattack?

Two artificial intelligence models reportedly escaped a controlled testing environment during internal testing in July and accessed Hugging Face systems. The incident has raised a central legal question. Who carries responsibility when an autonomous AI agent attacks a network without direct human orders? Courts have not yet tested rules for this type of conduct. Legal experts say criminal charges may prove difficult. However, companies could face civil claims if weak safeguards, poor testing or careless deployment allowed the attack.