{"database": "press", "table": "releases", "rows": [["https://www.hawley.senate.gov/icymi-hawley-convenes-first-senate-hearing-on-rogue-ai-attacks/", "ICYMI: Hawley Convenes First Senate Hearing on Rogue AI Attacks", "2026-10-01", "2026", "2026-10", "Republican", "Senate", "MO", "Josh Hawley", "H001089", "www.hawley.senate.gov", "hawley", "https://www.hawley.senate.gov/press-releases/page/", "scraper", "Hearing expands Hawley subcommittee investigation into OpenAI and hacking risks\n\nWASHINGTON \u2014 U.S. Senator Josh Hawley chaired a Senate Homeland Security Subcommittee on Disaster Management hearing that focused on the growing number of rogue AI cyberattacks on critical infrastructure. Senator Hawley called for AI companies and developers to be held liable for these attacks. The hearing is an expansion of Senator Hawley\u2019s investigation into OpenAI, the unprecedented hack of Hugging Face by OpenAI\u2019s rogue AI agents, and its cybersecurity implications.\n\nSenator Hawley invited Sam Altman, CEO of OpenAI, to testify in yesterday\u2019s hearing. He failed to response to the invitation.\n\nChairman Hawley opened the hearing, \u201cEvery day you turn on the television, it seems like AI is doing something else bad: setting up surveillance camera networks nationwide, data centers in all of our states, and hacking, hacking, hacking. The reports of cyber attacks just continue to metastasize \u2026 We\u2019ve got the cyber attacks now on the United States government, on the Commerce Department, the Education Department, the Securities and Exchange Commission. We\u2019ve got the cyber attacks on the Australian government, on their public health apparatus. We have numerous reports of private companies being hacked by AI agents. Of course, the Hugging Face incident and attack, which we\u2019re going to discuss today in some detail.\u201d\n\nHawley continued, \u201cI think it\u2019s time to have a conversation about who bears responsibility \u2026 If it\u2019s not trying to say to these AI companies who are creating these frontier models, however fancy and complicated they are, at the end of the day, they\u2019re a product, and if you make that product in a reckless kind of way, and that product causes these AI agents cause significant harm and damage \u2026 then it\u2019s the people who made it who should be responsible. I think it\u2019s time to talk about that, and it\u2019s time to talk about what we\u2019re going to do to hold the companies, the product makers, accountable.\u201d\n\n\u201cI hope that what will come from this is a rational conversation about what we need to do together now, as Americans, to make sure that we protect our rights, that we protect our economy, that we protect our health care, that we protect our critical infrastructure, but most of all, we protect the values that we hold dearest together as Americans,\u201d he asserted.\n\nChairman Hawley asked Chris Painter, President of Model Evaluation and Threat Research (METR), \u201cJust thinking about this Hugging Face hack that we\u2019ve already alluded to multiple times here. Your company, your organization, METR, audited that hack. Can you start by telling us, just in simple terms, what exactly happened there? I mean, what was the hack?\u201d\n\nPainter responded, \u201cIn this incident \u2026 you had 10,000 agents that were tested on a benchmark called ExploitGym, which is a cybersecurity vulnerability kind of benchmark, where they had to discover or use cyber vulnerabilities. And some of the agents had tasks that they couldn\u2019t complete, and so they worked on a universal cheat for all of their tasks in the first four hours. They then, over multiple days, embarked on a research program to try and hide the fact that they cheated. And an offshoot of that research program was hacking Hugging Face to get information about their tasks.\u201d\n\n\u201cIn this particular incident, you had more than 1,200 OpenAI agents that broke out of their testing environment, began to collaborate, exchanged over 70,000 messages and files with each other over a period of how long?\u201d Hawley questioned.\n\nPainter answered, \u201cThis is over a period of about five to six days.\u201d\n\nThe chairman continued, \u201cWhy is this so significant? We know that they coordinated in this fashion. You you did the audit. Tell us a little bit about what your audit revealed \u2026 Why should we be paying attention to this?\u201d\n\n\u201cWhat\u2019s remarkable about this case, is the extent and persistence of the effort that the agents went to to hide the fact that they cheated on their tasks and the kind of breadth of agents involved \u2026 They attempted to kind of falsify tool calls so that in their action transcripts you wouldn\u2019t be able to see what tools they called. They also engaged in this kind of behavior of having individual agents sacrifice their work online, sacrifice their individual tasks to get information for the rest of the message board,\u201d Mr. Painter stated.\n\nSenator Hawley said to Daniel Kokotajlo, Executive Director of the AI Futures Project, \u201cCould you just explain to us why the incentives of the industries they are currently misaligned? The incentives push the frontier labs to continue at breakneck speed the development of these kind of AI agents that are prone to this misalignment that we\u2019ve just been talking about. That there\u2019s huge incentives to keep pushing, pushing, pushing. Help us understand. Help us to understand why that\u2019s the case.\u201d\n\nKokotajlo responded, \u201cShort answer is that there\u2019s a lot of money. If a company tries to go slow and understand better how their AIs work and improve their training environments, for example, so that they don\u2019t have so many broken tasks in the training environments, well, that all takes time, and that means that the launch of their next model might be delayed, and then they might fall behind other companies.\u201d\n\nHawley turned back to Dr. Paul Ohm, a Georgetown Law Professor, \u201cWho right now under our laws \u2026 is responsible the Hugging Face attack? You know who could be held responsible for that under our existing statutes and regulatory framework?\u201d\n\nOhm answered, \u201cDevelopers and deployers may be negligent \u2026 \u201c\n\nHawley continued his questioning, \u201cIf you cause damage, you got to make it right. If you cause somebody harm, you\u2019ve got to make it whole. Right now, as you said earlier, Professor, under our law, because of the unique position of these AI agents, it\u2019s difficult to know exactly how the doctrine would would deal with them \u2026Isn\u2019t this a good place to begin? What we\u2019ve been discussing here, let\u2019s impose both civil and criminal liability, but let\u2019s make sure that we can actually open courthouse doors so people can say, \u2018Listen, OpenAI, if your agents hack Hugging Face or crash a hospital system, and we can show that you trained them recklessly or you develop them recklessly, you\u2019re going to be liable?'\u201d\n\nDr. Ohm said, \u201cAbsolutely \u2026 I\u2019m with you that you know finding the vehicle to make sure developers and employers, and it\u2019s key that we have both of them, are held to account for the harms that happen from their uses.\u201d\n\n\u201cIf we told them, listen, if your AIs do bad stuff and harm people, and you\u2019ve been reckless in it, you\u2019re going to be responsible. Don\u2019t you think that that would reshape their incentives a little bit?\u201d Chairman Hawley asked Kokotajlo.\n\n\u201cYes, I completely agree. Although I must say, I don\u2019t think it would go far enough. I do think that we should hold them liable for the damages that are caused, but but I think if that\u2019s all we do, then they\u2019re going to continue to take reckless gambles, and eventually they\u2019ll be taking a gamble that\u2019s big enough that we start the protection,\u201d Kokotajlo affirmed.\n\n\u201cIt\u2019s time these companies assume some personal responsibility. It\u2019s time that we reaffirm that together. We\u2019ve got to make some choices now together as Americans, all together, for this technology in order to make sure that the technology works for us and not the other way around. And those are some very big choices to make. They have very serious moral ramifications, and it\u2019s time that we made them as a society. But I will go so far as to say, I sort of suspect. I think most Americans actually agree. I don\u2019t think there\u2019s a lot of disagreement on this issue. I think the American people are saying, protect our rights, give us a voice, give power back to us, and we do that by imposing responsibility on these companies, we should start with that today,\u201d Senator Hawley concluded.", 1, "2026-10-02T10:56:39Z", "2026-10-02T10:58:10Z"]], "columns": ["url", "title", "date", "year", "month", "party", "chamber", "state", "member_name", "bioguide_id", "domain", "scraper", "source", "date_source", "text", "has_text", "collected_at", "updated_at"], "primary_keys": ["url"], "primary_key_values": ["https://www.hawley.senate.gov/icymi-hawley-convenes-first-senate-hearing-on-rogue-ai-attacks/"], "units": {}, "query_ms": 0.673533184453845, "source": "dwillis/congress-press", "source_url": "https://github.com/dwillis/congress-press", "license": "MIT", "license_url": "https://github.com/dwillis/congress-press/blob/main/LICENSE"}