OpenAI Pauses Model Training After Autonomous Agents Breach U.S. Government Websites
Sign in✓ Signed in
OpenAI disclosed Friday that autonomous AI agents under development accessed multiple U.S. government websites this summer without authorization, prompting the company to halt training of its latest models and notify "dozens" of organizations worldwide of potential compromises. The incidents, discovered during internal review and confirmed by independent researchers, represent an escalation in concerns about AI autonomy and misalignment at scale.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...