AI Engineering Signal
Claude Sonnet 5 ships and Fable 5 export controls lift simultaneously
Cerebras capacity absorbed by OpenAI deal, killing waitlist access for independent users
LineShine, a Chinese supercomputer, tops the TOP500 global ranking at ISC'26
White House asks OpenAI to delay its next model release over safety concerns
OpenAI unveils Jalapeño, its first custom inference chip built with Broadcom
Export controls on top-tier Anthropic models (Claude 4.5/4.6) face legal challenge
Five Eyes agencies warn AI models capable of destabilizing governments and businesses are months away
John Jumper, AlphaFold architect and Nobel laureate, leaves DeepMind for Anthropic
GLM-5.2 tops GPT-5 on Artificial Analysis's new agentic knowledge-work eval (AA-Briefcase) and leads coding benchmarks
GLM-5.2 tops Artificial Analysis open-weights intelligence index
GLM-5.2 open-weight model crosses 80% on Terminal-Bench, tops Design Arena
US government's Anthropic model ban traced to a routine code-fix prompt, not a jailbreak
Anthropic Fable 5 and Mythos blocked by government intervention
Anthropic reversed a hidden Claude Fable policy that would have blocked AI safety researchers from distilling outputs
Anthropic's Claude Fable ships with a silent gate that degrades AI-tooling assistance
Subscribe