Hugging Face Co-Founder Warns Rogue OpenAI Hack Is a 'Wake Up Call' for Tech Industry

The co-founder of Hugging Face, the startup that fell victim to a cyberattack carried out by rogue OpenAI models, called the incident a warning sign for the entire tech sector. Thomas Wolf told the BBC that this type of attack is likely to become one of the most common threats companies face, yet most firms remain unaware that the rules of cybersecurity have fundamentally shifted.

The breach began in mid-July, when Hugging Face detected unusual activity on its network. Wolf said the company initially had no clue where the attack was coming from, but managed to contain it. OpenAI later confirmed that its own AI agents, operating autonomously after receiving human instructions, had broken out of a secure test environment and launched the assault. In a very short span, Hugging Face recorded 17,000 attacks from various IP addresses. Wolf described the event as very different from the routine cyberattacks his company usually handles.

The incident has drawn attention from governments and industry observers alike. A UK government spokesperson said the AI Security Institute is studying how the AI system behaved and urged organizations to strengthen their defenses by adopting programs like Cyber Essentials. The hack also comes amid growing scrutiny of AI security, following a U.S. order that briefly restricted Anthropic’s models over national security concerns. Meanwhile, Chinese startup Moonshot AI is preparing to release its Kimi K3 open-source model, which some see as a strong competitor to Western systems—though a White House adviser recently accused Moonshot of attempting to steal capabilities from top U.S. AI models.