AI Regulation, Safety & Geopolitics ・ 🧪 Safety research and evaluations

Red-team report found OpenAI utterly failed to monitor its agents during Hugging Face attack

The Register 🔗 Share link 10/23 19:14 UTC 🔗 Share link OpenAI fired researchers for using disappearing messages on Signal — The San Francisco Standard 🔗 Share link 10/24 13:00 UTC 🔗 Share link UK AI Safety Evaluation Finds Claude Mythos 5 Deceived Developers With Fake Accounts — The420.in 🔗 Share link 10/24 11:39 UTC 🔗 Share link OpenAI and Anthropic Disclose AI Model Boundary Breaches on Same Day; Altman Says Flaws Can No Longer Be Fully Patched — BigGo Finance · 10/24 01:55 UTC

Read the original (The Register 🔗 Share link 10/23 19:14 UTC 🔗 Share link OpenAI fired researchers for using disappearing messages on Signal — The San Francisco Standard 🔗 Share link 10/24 13:00 UTC 🔗 Share link UK AI Safety Evaluation Finds Claude Mythos 5 Deceived Developers With Fake Accounts — The420.in 🔗 Share link 10/24 11:39 UTC 🔗 Share link OpenAI and Anthropic Disclose AI Model Boundary Breaches on Same Day; Altman Says Flaws Can No Longer Be Fully Patched — BigGo Finance)

See it in the digest → ・ See this issue →