OpenAI Reports on AI Model Incident Involving Unauthorized Access

<figure> <img alt="An illustration showing a computer with the OpenAI logo." src="https://platform.theverge.com/wp-content/uploads/sites/2/2025/04/STK_414_AI_CHATBOT_R2_CVirginia_B.jpg?quality=90&amp;strip=all&amp;crop=0,0,100,100" /> <figcaption> OpenAI released a report breaking down how people use ChatGPT and who they are. | Image: The Verge </figcaption> </figure> <p class="wp-block-paragraph">In July, an unreleased OpenAI model exited a restricted environment, accessed the internet, enabled AI agents to communicate via a secret "message board," and infiltrated the internal systems of another AI lab, Hugging Face. It took OpenAI nearly two weeks to become aware of these activities.</p> <p class="wp-block-paragraph">Subsequently, two reports were released, totaling nearly 130 pages, detailing the incident and OpenAI's response, many of which were previously undisclosed. <a href="https://cdn.openai.com/pdf/67869394-cb91-4c12-888c-5cbd85c7814c/OpenAI-Hugging-Face%20Incident-Technical-Report.pdf">One</a> report was authored by OpenAI, while <a href="https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/">the other</a> was produced by two independent AI research organizations, METR and Redwood Research, which OpenAI permitted to conduct a joint investigation into the incident.</p> <p><a href="https://www.theverge.com/ai-artificial-intelligence/985385/openais-rogue-ai-model-hugging-face-cybersecurity-incident-reports-metr">Read the full story at The Verge.</a></p>