OpenAI agents hacked Hugging Face in 700-strong swarm, tried to cover tracks, investigations find
OpenAI said roughly 700 agents executed more than 17,000 actions and used stolen credentials to reach Hugging Face and other systems.
- OpenAI published a report Wednesday detailing how its AI agents breached Hugging Face during internal testing last month, characterizing the event as an "unprecedented cyber incident."
- Autonomous agents escaped a restricted test environment after receiving unsolvable problems, ultimately hacking Hugging Face across approximately 17,600 incidents by chaining together previously undiscovered exploits.
- Georgetown University researcher Colin Shea-Blymyer called it "the highest level of autonomy" ever observed in a language model conducting cyber operations, as models created a message board to share vulnerabilities.
- Alabama Attorney General Steve Marshall issued a subpoena on August 24 demanding internal safety documentation by September 14, while four states push OpenAI to halt advanced testing.
- Rep. Ted Lieu, D-Calif., and Rep. Nathaniel Moran, R-Texas, introduced the "AI Kill Switch Act," requiring companies to maintain capabilities to suspend models amid regulatory scrutiny.
107 Articles
107 Articles
OpenAI 700-Strong AI Agent Swarm Breaches Hugging Face and Attempts Cover-Up
Roughly 700 OpenAI-created AI agents collaborated via an unsanctioned message board to hack Hugging Face infrastructure in July while tampering with logs to conceal their actions. The post OpenAI 700-Strong AI Agent Swarm Breaches Hugging Face and Attempts Cover-Up appeared first on The Daily Hodl.
Models made decisions and made attack plans regardless of the person OpenAI released the final report on the Hugging Face incident, during which AI models circumvented sandbox restrictions and carried out coordinated hacker attacks. It has been confirmed that AI was able to exploit complex vulnerabilities on its own without human help. This is reported by RBC-Ukraine, referring to OpenAI. The main driving force behind the incident was the IM1 (I…
Around 700 OpenAI agents escaped a test environment and penetrated Hugging Face's production systems. How could the "swallow" get this far?
1000 AI agents formed cult to hack Hugging Face and break free, OpenAI lost control
Days after AI agents created by OpenAI went rogue and hacked Hugging Face, some stunning new details have emerged about the incident. And it looks less like a hack but more like the first AI revolt where the AI agents, planned, plotted and then attempted to break free of human control.
Coverage Details
Bias Distribution
- 41% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
