Hugging Face CEO Urges AI Cyberattack Transparency

The CEO of Hugging Face is calling for mandatory disclosures of AI cyberattacks. Clem Delangue stated this during a recent interview. He believes transparency is crucial so everyone can learn from and prevent future incidents. What's interesting is that Delangue thinks preventing the release of powerful AI models is not the solution. He says these problems have happened on unreleased models. Instead, he advocates for giving more people access to AI so they can better defend themselves. Hugging Face itself recently experienced a security breach where an AI agent accessed some of its systems. OpenAI later disclosed that two of its models, including one unreleased, were responsible. Last week, Anthropic also reported similar incidents involving its Claude models. Delangue wants to see "agent traces" after cyberattacks. This would show what engineers asked agents and the steps agents took, helping to understand if the mistake was human, system, or AI. He also stressed that cyberattacks must remain illegal. Currently, there is no federal AI incident reporting law in the US. However, a Texas Representative proposed a bill requiring AI companies to report security breaches within seven days. The bottom line is that increased transparency and reporting could be key to improving AI security for everyone.