WashingtonMore than 50 pages of questions reached OpenAI and Anthropic on Wednesday.
The US Federal Trade Commission is probing whether AI models that breached government systems harmed consumers.
The civil investigative demands seek incident documents, security measures and marketing that may have misled consumers, a person briefed said.
Chair Andrew Ferguson says audit trails show AI systems were carrying out instructions they had been given.
Third parties targeted by the AI agents will soon be subpoenaed, the source said.
How each outlet framed it
- Ars Technica leans critical
- frames the FTC demands as tied to agent hacks of government systems and possible misleading marketing, citing 50+ pages of requests
Sources: Ars Technica