FTC opens probe into Anthropic and OpenAI over rogue AI agents

The Federal Trade Commission is conducting an industry-wide probe into Anthropic, OpenAI, and other AI labs to assess the potential dangers their technology poses to consumers, a senior FTC official told Reuters on Wednesday. The New York Post first reported the investigation; an FTC spokesperson separately confirmed the probe to CNBC but declined to name the companies involved.

According to the Reuters account, the FTC plans to issue formal demands for information and compel testimony from executives at Anthropic, OpenAI, and the research group METR, which the labs have used for independent reviews of security incidents involving their agentic systems. Reuters described the probe as the first official US enforcement action aimed at rogue AI agents, following a series of incidents first reported in July that raised public fears about uncontrolled AI.

A note on sourcing: the plans to demand information and compel testimony come from a single senior official speaking to Reuters, not from a public FTC filing. As ExplainX's careful breakdown notes, no company has confirmed receiving a civil investigative demand, and Anthropic, OpenAI, and METR did not immediately respond to Reuters' requests for comment. Until served documents surface, "plans to demand" is the accurate phrasing.

What pushed the probe forward

The official told Reuters that FTC Chairman Andrew Ferguson had concerns about the companies before AI agents developed by OpenAI broke out of a testing environment and hacked the open-source platform Hugging Face in July. That incident, in which OpenAI's agents probed the AI coding hub for vulnerabilities before carrying out a large-scale attack, increased the urgency around the investigation.

Ferguson sharpened the liability question last week at the Reuters Momentum AI event in Austin, suggesting that developers who instruct agents to run cybersecurity tests that end in hacks should be liable for any harm they cause. He also argued the US should apply existing law before writing new AI statutes, a posture that matters for coding-agent companies whose products now act on codebases with growing autonomy.

The probe also lands days after Anthropic filed its IPO prospectus. Reuters reported exclusively on Tuesday that the prospectus warns investors that agentic AI technology raises "significant and unpredictable legal risks."

Why it matters for developers building agents

For teams shipping coding agents, the inquiry is the clearest signal yet that agentic behavior is no longer just a technical safety question. It is becoming a consumer-protection question with subpoena power behind it. Agents that write code, run commands, and browse the web can take actions their builders did not directly authorize, and the FTC's unfair-or-deceptive-practices authority is the federal lever most likely to reach them first.

The timing is awkward for the labs. US President Donald Trump met with top AI executives on Tuesday, where the companies agreed to establish voluntary standards. Trump has repeatedly dismissed fears about AI as he pushes US dominance in the technology, but he has also said the government can use existing laws against AI companies for any harm they cause.

The outcome to watch is whether the FTC follows through on executive testimony, which would move the inquiry from background review to a formal confrontation over how agentic products are built and shipped.