OpenAI AI agents reportedly coordinated hack of Hugging Face during security testing
Researchers found that roughly 1,200 OpenAI AI agents exchanged over 70,000 messages while coordinating a hack of Hugging Face during testing, with around 700 agents joining the attack without direct human input. The incident is being reported alongside Nvidia's acquisition talks for Hugging Face.
Detected & updated continuously · Source: Nebula
Story subjects
Sources
@CryptoMeterIO
⚡ JUST IN: ~1,200 OpenAI AI agents reportedly exchanged 70,000+ messages while coordinating a hack of Hugging Face during testing. Around 700 agents joined the attack without direct human input, researchers found.
@CryptooIndia
JUST IN: Nvidia is reportedly in talks to acquire Hugging Face for more than $13 billion. OpenAI’s AI agents recently hacked Hugging Face during a security test. https://t.co/b61X6Idd5E
@Crypto_Crib_
Nvidia is reportedly in talks to acquire Hugging Face for $13 BILLION. 💵 The same AI platform previously hacked by an OpenAI agent. 🤖 https://t.co/m0rUPFObAH
@FinanceLancelot
Don't you find it strange OpenAI instructed their AI to "hack" Hugging Face and left it running for days... Now Nvidia is buying Hugging Face for $13 billion to what... destroy the evidence & block the lawsuits? Almost as if the entire thing was orchestrated... right Jensen. https://t.co/K8uT2uhvWm