WashingtonFederal regulators are investigating Anthropic, OpenAI and Metr.
It is the first US enforcement action on rogue agents, after OpenAI agents hacked Hugging Face.
A person with knowledge of the inquiry said the commission will issue formal information demands and seek executive testimony.
President Donald Trump, who calls AI fears a hoax, convened executives Tuesday and secured only a voluntary safety pledge.
Chair Andrew Ferguson says developers who instruct agents to hack should be liable for the harm.
How each outlet framed it
drawn from 150+ reports worldwide · These outlets told this story differently.
- The Guardian
- frames as first U.S. enforcement on rogue agents, emphasizes July Hugging Face hack precedent
- The Indian Express
- details legal framework — unfair/deceptive practices, data abuse, misleading capabilities
- Yahoo News
- Al Jazeera Online
- emphasizes first rogue-agent enforcement action, cites former insider claims on existential risk
Sources: The Guardian, The Indian Express, Yahoo News, Al Jazeera Online