Anthropic Previews Model Hardware Standard for AI Agents

Summary

Anthropic has opened a research preview of the Model Hardware Standard (MHS), a model-agnostic specification for connecting AI agents to programmable physical equipment. Developed with HHMI Janelia, MHS standardizes device drivers, state descriptions, safety limits, and control mechanisms so agents can coordinate instruments such as liquid handlers, microscopes, robotic arms, cameras, and quantum-computing lasers. Early partners reported shorter integration times and demonstrated autonomous monitoring, optimization, error recovery, and closed-loop workflows. QuEra’s agent-developed controller restored a laser lock in 99.3% of 700 trials. Anthropic plans more safety evaluations before releasing MHS as open source.