Mike: open-source legal AI
Also: Cisco releases open-source toolkit for verifyingAI modellineage - Help Net Security
Shai-Hulud Themed Malware Found in the PyTorch Lightning AI Training Library
The Zig project's rationale for their firm anti-AI contribution policy
Also: The Zig project's rationale for their firm anti-AI contribution policy
Claude Code refuses requests or charges extra if your commits mention "OpenClaw"
Also: Claude Code refuses requests or charges extra if your commits mention "OpenClaw"
DataCenter.FM – background noise app featuring the sound of the AI bubble
Finetuning Activates Verbatim Recall of Copyrighted Books in LLMs
Also: Alignment whack-a-mole: Finetuning activates recall of copyrighted books in LLMs
OpenAI announces new advanced security for ChatGPT accounts, including a partnership with Yubico
Also: OpenAI’s new security model is for ‘critical cyber defenders’ only • OpenAI announces new advanced security for ChatGPT accounts, including a partnership with Yubico
GPT-5.5 Tops Academic Benchmarks but Loses to Rivals in Real-User Tests
Also: OpenAI Codex 系统提示词披露，GPT-5.5 AI 模型“永不谈论哥布林” - IT之家 • GPT-5.5 Tops Academic Benchmarks but Loses to Rivals in Real-User Tests
Just Now:GPT-5.6 Exposed,GPT-5.5 Goblin Obsession Leads to Overnight Block by OpenAI - 36氪
Also: ChatGPT Developed A Goblin Obsession After OpenAI Tried To Make It Nerdy - Engadget • GPT-5.5-Cyber Announced: OpenAI CEO Sam Altman Rolls Out Specialised Frontier Model ...
This AI Actually Surprised Me
Also: ClaudeSecurity Just Launched — AI Code Scanning That Actually Reasons (Public Beta)
AISI EvaluatesGPT-5.5 Cybersecurity Performance Against Advanced Tasks
Also: OpenAI LaunchesGPT-5.5-Cyber Model for Cybersecurity Defense | Intellectia.AI • AISI EvaluatesGPT-5.5 Cybersecurity Performance Against Advanced Tasks
Chrome looks set to ship an LLM Prompt API to the web. We oppose this API
Also: Chrome looks set to ship an LLM Prompt API to the web. We oppose this API