Ai Research

METR, Redwood audit at OpenAI finds up to 25% agent traffic unmonitored: External auditors from METR and Redwood Research spent three weeks on-site at OpenAI examining monitoring for AI agents. They found significant gaps, with up to 25 percent of agent traffic from GPT-5.5-class and above models going unmonitored.

AI Safety • The Decoder

Coding-agent harness swaps barely change pass rates but can triple cost, study finds: A controlled study found that switching coding-agent harnesses such as Claude Code, OpenCode or mini-SWE-agent barely changed pass rates versus rerunning the same setup, while cost per task could triple. The higher cost was driven mainly by system prompts resent at every step.

AI Research • The Decoder

OpenAI built an offline web world to train isolated models to browse and use apps: OpenAI said it built an “offline world” from cached web content to train isolated frontier models to browse, search for information, and use web apps. It described computing environments where models can write and run code.

AI Research • Techmeme

Anthropic and Citrini use Claude to design scorpion toxin minibinders, test 211: Anthropic and Citrini Research used an internal Claude model to autonomously generate protein minibinders against five scorpion neurotoxins and tested 211 designs. Anthropic said it had previously shown Claude designing minibinders from a single prompt in August.

AI for Science • Techmeme

Meta pauses Muse sharing after security firm reportedly found over 6,800 links: Meta said it paused a Muse feature that lets users share conversations with its AI assistant after researchers found the URLs publicly available on the web. Security firm Claroc said it found more than 6,800 Muse links through public DNS records, Ars Technica reported.

AI Privacy • The Verge AI

A Potentially Historic AI Breakthrough: A video examines Nick Cammarata's essay arguing neural networks may be more understandable than researchers believed, with apparent complexity possibly reflecting how models are examined. It links cat-face and car-front representations and considers detecting AI deception; no evidence is described.

AI Research • The AI Daily Brief: Artificial Intelligence News