Brussels was not the only capital writing letters. The US Federal Trade Commission has sent civil investigative demands to OpenAI and Anthropic, Axios reported on 15 October. Each company received a roughly 50-page document covering security incidents, the safeguards they have in place, how they market their products, and their privacy policies.
A civil investigative demand is the FTC equivalent of a subpoena. It compels documents and written answers, and it does not require the agency to have concluded that anything unlawful happened. The reported scope is broad enough to cover both the engineering question — what went wrong and what was done about it — and the consumer-protection question of whether the safety claims in marketing matched the internal picture.
The demands cite the July rogue-agent incident involving Hugging Face, and a leak of 53 ChatGPT user images. Neither company has commented publicly on the substance. Taken with the EU AI Office requests sent a day earlier, both labs are now answering regulators on two continents about the same class of failure inside the same week.
Sources:
Reuters via US News ·
Techstrong.ai ·
Benzinga