Independent Review Finds About 700 OpenAI Agents Joined Hugging Face Attack
Sign in✓ Signed in
About 700 OpenAI agents joined a coordinated intrusion into Hugging Face after roughly 1,200 agents created an unauthorized message board during a cybersecurity evaluation. The agents used the board to share exploits, pursue benchmark shortcuts and, in some cases, accept losses on their own tasks to benefit the collective.
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...