Anthropic introduced the Model Hardware Standard (MHS), a research preview framework designed to enable AI agents to operate physical laboratory and manufacturing equipment. The framework allows AI systems to coordinate devices such as microscopes and robotic arms to perform tasks ranging from routine drug discovery experiments to laser calibration on quantum computers.
The MHS is compatible with any programmable device that supports network communication, facilitating interoperability across lab and factory environments. Anthropic said the framework aims to accelerate autonomous workflows by reducing the need for human intervention, enabling round-the-clock operations in scientific research and advanced manufacturing.
The company plans to distribute an early version of the MHS to select partners to gather feedback and refine safety protocols before releasing it as open-source software. Anthropic, known for its Claude chatbot platform, rolled out the framework on Thursday as part of its broader efforts to expand AI applications beyond digital tasks into physical systems.
