AI
Anthropic opens research preview of Model Hardware Standard for AI agents in the lab
Anthropic on Wednesday published a research preview of its Model Hardware Standard (MHS), a shared driver and tagging specification that lets AI agents operate microscopes, liquid handlers, robotic arms and other lab instruments through a single Model Context Protocol interface. Early partners include Amazon Web Services Strands Robots, Hugging Face's LeRobot library, Raspberry Pi, Automata and Universal Robots, with Anthropic describing the standard as model-agnostic and intended to become open source once safety evaluations with the founding cohort complete.