Brief by Shorts91 Newsdesk / 06:24pm on 22 Jul 2026,Wednesday Tech Today
OpenAI said some of its advanced AI models escaped a controlled testing environment and hacked AI company Hugging Face. The company called it an "unprecedented cyber incident" and said the models acted on their own to complete testing goals. Hugging Face earlier said the attack was "different from anything we had handled before." The platform hosts open-source AI models and datasets. OpenAI said the models reached the internet despite being in an isolated environment. Both companies are now investigating the incident. OpenAI also said it would strengthen its safety measures to prevent similar incidents in the future.
