FTC intensifies OpenAI, Anthropic probe following 'rogue AI' attacks

Sept. 30, 2026, 3:43 p.m. ET

Does artificial intelligence pose a real threat to the American public?

A potentially far-reaching Federal Trade Commission probe into Anthropic, OpenAI, and other top AI labs is shifting into high gear to find out.

According to a senior FTC official who spoke to USA TODAY on the condition of anonymity, the agency launched the investigation before the rash of rogue agent incidents but plans in coming weeks to intensify that inquiry, using civil investigative demands – similar to subpoenas – to obtain more information and compel executives to testify.

The probe is the first official enforcement action to target the actions of rogue agents. The July "Hugging Face incident" and subsequent breaches sparked fears that AI could harm humans.

"This is a sweeping and significant probe," the senior FTC official told USA TODAY.

Specifically, the probe will turn on whether the companies engaged in deceptive or unfair business practices that violate the FTC Act, the federal law enacted in 1914 that established the agency and gave it broad powers to protect consumers and maintain fair business competition. The FTC has used this authority in the past against companies that did not secure consumers’ data.