Anthropic has opened a research preview of a new specification that lets AI agents directly control laboratory hardware and manufacturing equipment. The Model Hardware Standard (MHS) aims to cut the time between hypothesis and physical experiment by giving AI models a standardized way to operate devices like microscopes, liquid handlers, and robotic arms.
What MHS does
The specification creates a shared driver layer that allows AI agents to discover, connect to, and operate programmable scientific instruments. Instead of spending weeks writing custom integration code for each piece of lab equipment, researchers can use MHS-compatible devices that Claude agents control out of the box. The system supports parallel operation across multiple instruments and enforces device-level safety boundaries to prevent hardware damage.
"The conceptual leap here is going from Claude agents working in the digital world [of] analysis, hypothesis generation . . . to do experiments in the physical world, in the lab," Jonah Cool, who works within Anthropic to help develop its life sciences strategy, told the Financial Times.
Early trials show practical results. QuEra ran automated quantum laser-lock recovery experiments with high success rates. Carnegie Mellon University executed autonomous lab experiments with minimal human intervention. The standardized driver layers reduced hardware setup times from weeks to hours.
How partners are adopting it
Several hardware and software companies are building MHS support into their equipment. AWS will provide a private, pre-release version of its Strands Robots library for connecting AI agents to physical devices during the preview. Automata is adding MHS to its LINQ lab automation platform for intelligent error handling. Doosan Robotics is testing the specification with robotic arms for automated quality assurance and multi-robot coordination.
MBF Bioscience is building an MHS driver for ScanImage, the software that runs laser-scanning microscopes in hundreds of neuroscience labs worldwide. QIAGEN has a working proof-of-concept on its QIAsymphony Connect nucleic acid purification platform, showing how AI agents could help labs troubleshoot instrument issues faster and improve uptime. Danaher and Anthropic are exploring how MHS could scale biomedical research and development across smart instruments and autonomous laboratories.
Anthropic's broader research push
The MHS launch follows Anthropic's June release of Claude Science, a product that integrates common research tools and provides flexible access to computing resources. The company has also expanded its AI for Science program, which offers free credits to researchers working on high-impact scientific projects. Eric Kauderer-Abrams, Head of Life Sciences at Anthropic, wrote on LinkedIn that the company is now broadening its focus "to include every scientific discipline, from physics, chemistry, math, materials science and beyond."
Anthropic plans to open-source the MHS specification after the preview safety testing period concludes. The company is currently preparing for an IPO expected in October, with some investors anticipating a valuation above $2 trillion, according to the Financial Times. The firm's annual revenue run rate topped $65 billion by the end of July, a person familiar with the matter told Reuters.
Why this matters for researchers
For scientists and lab managers, MHS addresses a concrete bottleneck: the manual translation between AI-generated experimental designs and physical execution. If the specification gains adoption across equipment vendors, it could reduce the integration work that currently separates computational analysis from wet-lab work. The open-source commitment after the safety preview means research labs won't be locked into a single vendor's stack. For those building skills in this area, the AI Learning Path for Research Scientists covers the intersection of AI agents and experimental workflows. The broader trend in AI for Science & Research is moving toward systems that don't just analyze data but act on it.
Your membership also unlocks:
