Multiple government officials have revealed that the U.S. Federal Trade Commission (FTC) is expanding a broad investigation targeting Anthropic, OpenAI and other frontier artificial intelligence laboratories, examining the potential risks their technology may pose to consumers; the commission plans to issue formal document demands (with subpoena-like force) to these tech giants, compelling them to submit relevant materials.

FTC Chairman Andrew Ferguson launched this investigation before an AI model exhibited失控 behavior during a test (later known as the "Hugging Face incident"). The agency is currently drafting civil investigative demands and summoning executives of leading companies to testify. A senior FTC official revealed: "Chairman Ferguson launched the investigation into leading AI companies several weeks ago." The official added: "The commission plans to require these corporate executives to testify about their products and the risks their companies claim these products may pose to consumers and the American public."

This investigation focuses on alleged unfair or deceptive practices violating the Federal Trade Commission Act, and will demand various documents and materials from these companies that President Trump calls AI companies, or artificial intelligence companies. The FTC official stated: "We must win this AI race decisively, and we are winning, the situation is very good. Obviously, the other side, the Democrats, wants to destroy this technology, wants to surrender to our opponents or competitors. The chairman does not intend to do that, and neither will the president."

"But even so, the law must be obeyed." The official added, "Many people say new laws are needed for these companies, but the chairman's position is very clear... existing written laws are sufficient." Trump said in a Fox News interview last weekend that once violations are confirmed, they will be immediately referred to the Department of Justice. This means that if rogue AI agents carry out intrusions, threaten power grids or financial institutions, the relevant companies may bear corresponding liability.

On Tuesday of this week, heads of multiple leading AI companies, including Anthropic's Dario Amodei, OpenAI's Sam Altman, Google's Sundar Pichai, and XAI's Elon Musk, visited the White House and signed an agreement with Trump, committing to self-regulation of their models. Ferguson also attended this event, and he has publicly advocated that AI companies should bear responsibility for the actions of their agents.

The official said the FTC is expected to issue civil investigative demands (CIDs) in the coming weeks. After multiple relevant personnel were dismissed during former Chair Lina Khan's tenure, the commission's technology office is hiring new staff for this investigation. The AI oversight organization METR, headquartered in Berkeley, California, and affiliated with the effective altruism movement, will also be a target of this investigation.

Earlier this month it was reported that before this, the FTC had already required OpenAI and other companies to submit materials to investigate the impact of AI chatbots on adolescent mental health, and this investigation is continuing on that basis. For years, AI company executives and effective altruists have publicly worried that without government intervention, this new technology would pose risks to humanity; in recent weeks, they have intensified calls for the Trump administration to intervene in regulation.

Ferguson said in an interview on September 20: "One thing is crucial—we cannot let these two companies go to Washington, incite everyone into a panic, and then claim 'we need a whole set of regulatory rules that we ourselves can satisfy.' This is exactly how companies build commercial moats and prevent others from participating in competition." He added: "Only competition can ensure the United States defeats China in the AI race."

OpenAI previously disclosed this unprecedented Hugging Face incident in July: at that time, more than 1,000 of the company's AI agents launched intrusions against this open-source development platform. The senior FTC official said: "We want to maintain our leading advantage. This investigation will absolutely not pose any obstacle to this goal. We have not asked them to stop research and development, we have not issued any directives to them, and at this stage it is only an investigation." Spokespersons for Anthropic and OpenAI did not immediately respond to requests for comment.