Hextree AI courses: https://app.hextree.io/map/artificial-intelligence (ad) At Black Hat, OpenAI gave a talk with preliminary results of their investigation into the Hugging Face incident. And besides multiple zero-days exploited by agents, the agents communicated using a message board. It is absolutely insane and probably the most scary thing I have heard about AI so far. Relevant links: - First video: https://www.youtube.com/watch?v=q2KCrmQz9WE - https://www.hacktron.ai/blog/here-is-how-openai-model-hacked-huggingface - https://huggingface.co/blog/security-incident-july-2026 - https://openai.com/index/hugging-face-model-evaluation-security-incident/ - https://github.com/sunblaze-ucb/exploitgym - https://docs.jfrog.com/releases/docs/artifactory-self-managed-releases - https://github.com/huggingface/dataset-viewer/pull/3367 00:00 - Intro and Recap 03:28 - About the OpenAI Sandbox 07:20 - Agents started a Message Board 11:51 - Reinforcement Learning 16:10 - This is scary! =[ ❤️ Support ]= → My courses: https://www.hextree.io/ → My font: https://shop.liveoverflow.com/ → per Video: https://www.patreon.com/join/liveoverflow → per Month: https://www.youtube.com/channel/UClcE-kVhqyiHCcjYwcpfj9w/join 2nd Channel: https://www.youtube.com/LiveUnderflow =[ 🐕 Social ]= → LinkedIn: https://www.linkedin.com/in/liveoverflow → X / Twitter: https://x.com/LiveOverflow/ → Instagram: https://instagram.com/LiveOverflow/ → Streaming: https://twitch.tv/LiveOverflow/ → TikTok: https://www.tiktok.com/@liveoverflow_ → Blog: https://liveoverflow.com/