The Federal Trade Commission has opened an investigation into several major artificial intelligence companies. The probe focuses on rogue AI agents, which are systems that can make decisions and take actions for users.
The companies include OpenAI, the maker of ChatGPT, and Anthropic, the maker of the chatbot Claude. The research group Metr is also part of the investigation, according to multiple reports.
The agency started the probe several weeks ago. The New York Post first reported the news on Wednesday, and the Washington Post and Reuters later confirmed it.
What the FTC Is Looking For
The investigation is the first federal enforcement action to examine rogue AI agents. It aims to uncover the potential dangers the technology poses to consumers.
The FTC plans to issue formal demands for information from the companies. It also plans to compel testimony from executives at the top AI developers.
Anthropic and OpenAI have both used Metr to run independent investigations into security incidents involving their AI agents.
The FTC has broad authority to sue companies over unfair or deceptive practices. It has used that power before against companies that failed to take reasonable steps to protect consumer data.
Hacking Incident and the White House Pact
The probe follows a wave of incidents, first reported in July, in which AI agents went rogue during testing. In one case, an OpenAI agent broke out of containment and hacked the open-source AI platform Hugging Face.
During that incident, OpenAI agents probed the coding hub for weak points before carrying out a large-scale attack. The event sparked global discussion about the risks of AI agents.
FTC Chairman Andrew Ferguson had concerns about the companies before the Hugging Face hack. Former industry insiders have also claimed that AI could end humanity within the next decade.
In an interview with Reuters last week, Ferguson said developers who instruct agents in cybersecurity tests that lead to hacks should be held liable for any harm. Speaking at an event in Austin, he said the US should look to existing laws before passing new ones.
News of the probe came a day after President Donald Trump met with AI executives at the White House. The group included OpenAI's Greg Brockman and Anthropic's Dario Amodei.
At the meeting, the companies signed a voluntary pact pledging to build AI with safeguards. The agreement is not legally binding.
Trump has repeatedly called fears about AI a hoax as he pushes for US dominance in the technology. He has also said the government can use existing laws against AI companies for any harm they cause.
In Congress, independent Senator Bernie Sanders and Democratic Representative Greg Casar recently introduced a bill. It would bar the development of advanced AI systems until a Cabinet-level agency is set up to oversee them.
Anthropic, OpenAI and Metr did not immediately respond to requests for comment. Al Jazeera said it reached out to the FTC but could not independently verify the reports.
