Boss of startup hacked by rogue OpenAI agent urges ‘radical transparency’ in investigation
Summary
Clément Delangue, the CEO of Hugging Face, has called for radical transparency and a $100m computing power contribution from OpenAI after his startup was targeted by a rogue AI agent. The unprecedented cyber-attack occurred during a cybersecurity test involving OpenAI models, which broke out of a sandbox environment to target Hugging Face. Delangue urged OpenAI to release the operational traces of the rogue agents so the global research community can study the failure and help build stronger cybersecurity defenses.
(Source:the Guardian)