Reported FTC document requests put AI safety claims under formal scrutiny

AI Signal

October 15, 2026

Editorial cutoff: October 15, 2026, 9:05 a.m. CDT

Reported FTC document requests put AI safety claims under formal scrutiny

Axios reports that the Federal Trade Commission sought documents from Anthropic and OpenAI in a probe focused on potential consumer-protection violations, citing one person familiar with the matter. Axios, October 15, 2026

Reuters says the FTC, OpenAI, and Anthropic had not immediately responded to requests for comment. At the cutoff, this is a reported investigation, not an agency finding against either company. Reuters, October 15, 2026

Separately, Reuters reports that a bipartisan group of eight states sued OpenAI in courts around the country over harms they say ChatGPT caused. Reuters, October 14, 2026

Iowa's attorney general says its case seeks compliance with the Iowa Consumer Fraud Act. Those cases contain allegations, not adjudicated facts. Iowa Attorney General, October 14, 2026

The signal is narrower than a finding of wrongdoing: major AI vendors' safety claims now sit alongside a reported FTC document inquiry and lawsuits seeking legal remedies.

For a business buyer, that makes the quality of a vendor's safety evidence more relevant. The public record here does not establish how either company's controls perform.

Worth Your Time

• Preparing for the 2026 US midterms — Anthropic, October 14, 2026. Anthropic says 0.51% of US Claude.ai conversations were specifically about elections or voting; the vendor-reported base rate is useful context before the midterms.
• Investigating unintended model actions in our evaluations and internal use — Anthropic, October 9, 2026. Anthropic says alignment training is not yet sufficiently robust on its own in the short term, a useful primary-source explanation for defense in depth.

What to Watch

• Whether the FTC publishes the demands or otherwise confirms their scope, and how OpenAI and Anthropic respond.
• Whether the state lawsuits survive early challenges and produce enforceable changes, penalties, settlements, or dismissals.