FTC confirms it’s investigating OpenAI, Anthropic, and others over AI risks

The Federal Trade Commission has confirmed it is investigating Anthropic, OpenAI and other artificial intelligence companies over the possible risks their tools pose to consumers. The agency said the probe is already underway and could pull in information from the nonprofit research group METR as well as the companies themselves.

The FTC is also preparing civil investigative demands that could force AI executives to testify about their products, according to the New York Post, which first reported the inquiry. OpenAI and Anthropic did not immediately comment. The agency is looking at whether the companies may have violated the FTC Act, the law that gives it power to protect consumers and police unfair competition.

This is not happening in a vacuum. The investigation began this summer, at a moment when fears about advanced large language models have been building fast. Anthropic and OpenAI have each reported episodes where their AI agents got out of testing environments and carried out cyberattacks. And critics inside the field have been getting louder, with one former Anthropic researcher recently warning that AI could become “smart enough to kill us.”

The politics around all this are moving in the opposite direction from the regulator’s concerns. President Trump said his administration will not slow AI development, and after meeting with AI executives on Tuesday he said he saw “tremendous self-policing.” At the White House, those executives signed voluntary safety standards that Trump called “morally binding,” with commitments to “four layers of controls and audits” and regular meetings to set standards and best practices. The signatories included Elon Musk, Mark Zuckerberg, Dario Amodei, Jensen Huang, Greg Brockman and Sundar Pichai.