Research
LoC
METR Reports Accelerating AI Capabilities Progress at Anthropic
METR reports that AI-driven automation of R&D is accelerating Anthropic's capabilities progress, prompting some employees to significantly shorten their timelines for superintelligence.
XNov 6, 2026
Industry
LoC
OpenAI Introduces Kill Switch to Halt AI Research Models Within 35 Seconds
OpenAI reportedly deployed a safety mechanism that blocks a research model's outputs within 35 seconds of a human or automated monitor stopping its job, strengthening oversight of autonomous AI systems.
FirstpostNov 11, 2026
Research
LoC
Researchers Show GPT-6 Astra Can Improve AI Models by Directly Editing Their Weights
MIT researchers reported that Astra improved Qwen's math accuracy from 7.9% to 48.1% by modifying just 0.016% of its weights, demonstrating a potential new approach to AI self-improvement.
MITNov 6, 2026