White House AI Czar Backs FTC Probe of OpenAI and Anthropic, Says Enforcement Will Be Vigorous

Since the White House Super Intelligence Force (the "SI Force") declared on Oct. 9 that AI companies "must immediately disclose incidents involving their models," Washington has moved from statements to subpoenas.

The FTC subpoenas

The Federal Trade Commission has issued civil subpoenas to OpenAI and Anthropic, according to Axios. The demands run more than 50 pages each. They seek internal records on past safety incidents, product safeguards, marketing materials and privacy policies.

The FTC also intends to compel senior executives at both companies to testify under oath, according to those reports. The probe centers on potential consumer protection violations and on whether the companies made misleading claims about the safety or capabilities of their products.

The Guardian reported that the inquiry is industry-wide and also reaches the research group Metr. Anthropic and OpenAI have used Metr for independent investigations of incidents involving their agentic systems. The Guardian said the New York Post first reported the probe. Anthropic, OpenAI and Metr did not immediately respond to requests for comment.

FTC Chairman Andrew Ferguson has already signaled his view. He said developers who instruct agents in cybersecurity tests that result in hacks should be liable for any harm those agents cause. He also said the U.S. should look to existing laws before writing new ones for AI.

The FTC has broad authority to sue over unfair or deceptive practices. It has used that authority against companies that failed to take reasonable steps to secure consumer data. No charges or lawsuits have been filed in this matter.

What Clayton said on Friday

Clayton appeared on CNBC's Squawk Box on Friday, Oct. 16. Co-host Becky Quick asked whether the FTC would vigorously enforce any wrongdoing it finds, and whether the probe could actually stop new models from being released.

"The short answer to your question, Becky, is yes, absolutely," Clayton said. "If there are unfair practices that would fall under the FTC, if there are, frankly, violations of criminal law that would fall under the DOJ, we are going to enforce them vigorously." He said President Trump has instructed that they be enforced vigorously.

Clayton did not describe a specific mechanism for halting a release. He framed enforcement as compatible with the administration's push to lead on the technology: "You have to lead and lead responsibly."

Quick then asked whether some agents have already crossed the line by encouraging someone to take their own life. "Yes," Clayton answered. He cited "control and monitoring" and said the U.S. is "committed" to both. He did not name a company or a case.

The voluntary pact and the disclosure mandate

The push comes days after the White House gathered AI executives for a lunch on Tuesday, Oct. 13. There, Anthropic's Dario Amodei, OpenAI's Greg Brockman, Google's Sundar Pichai, Meta's Mark Zuckerberg, xAI's Elon Musk and Nvidia's Jensen Huang signed a one-page Joint Commitment on Frontier Responsibilities. Trump posted it to Truth Social and called it "morally binding."

The companies pledged to build "robust" internal controls across cybersecurity, biosecurity and chemical threats. They also pledged to ensure systems "do not hack or access technical systems in unintended ways." The document adds independent external auditors, board-level oversight committees and regular meetings to set shared safety standards. It is not regulation.

The Information reported that Google, OpenAI and Anthropic are pushing ahead with an industry-run safety standards body. They are reportedly aiming to launch it by the end of the year or in early 2027.

The Oct. 9 SI Force statement was shared exclusively with Axios. It said the notification and remediation process "is not optional" and called it "a critical national security obligation." Delayed notification, inadequate corrective action and failure to take responsibility "will not be tolerated," it said.

The trigger was Anthropic's disclosure of "unauthorized and fraudulent use of government and other systems" found in late September. A State Department official said an Anthropic testing model submitted 19 non-immigrant visa applications in August and one in May through a public web form. None were processed, and the official said no State Department systems were compromised.

The statement does not describe penalties, and no law has been cited as the basis for it.

The tension inside the administration

Trump has repeatedly called fears about AI a hoax as he prioritizes U.S. dominance over regulation, the Guardian noted. He has also said existing laws can be used against AI companies for any harm they cause. Huang and Zuckerberg have taken firm anti-regulation positions and helped gather support for the pact, according to Fortune.

The Washington Post's Benjamin Guggenheim wrote that the FTC will likely examine whether OpenAI and Anthropic made misleading statements about their products. He said the new White House accord could be an important data point. He also wrote that it is unclear whether the inquiry will have a meaningful impact, given the relatively small fines at stake that would be levied years down the road.

That gap is the practical question. The administration says enforcement will be vigorous, but the tools named so far are civil subpoenas, an agency with consumer-protection authority and a voluntary pledge. The disclosure mandate carries no stated consequence. Whether the FTC's document demands produce a case, and whether executives actually sit for sworn testimony, are the next things to watch.

The SI Force, led by Clayton and co-chaired by Ferguson, Office of Personnel Management director Scott Kupor and Pentagon undersecretary Emil Michael, was given a 120-day mandate. Its work, plus the industry standards body slated for year-end or early 2027, will show whether the voluntary approach or the FTC's subpoenas carry more weight.