Skip to content

Anthropic

Research preview opens for agent-to-machine control

Written by an AI model from the linked sources and checked against them by a second model. How we work

Video: twitter.com

Anthropic has opened the first phase of a research preview for the Model Hardware Standard (MHS), a standard meant to let AI agents safely operate physical equipment in scientific research and advanced manufacturing. The company announced it in an August 27 post on Twitter.

CNBC reported that MHS is designed to work with any device that has a programmable interface, and that Anthropic plans to open source it eventually. For now it is still only a preview, and the open-source release has no date.

Anthropic has done this before with software. It launched the Model Context Protocol (MCP) as an open standard on November 25, 2024, to replace the custom connector developers had to write for every pairing of AI system and tool. OpenAI, Google DeepMind and Microsoft adopted MCP in early 2025, and on December 9, 2025, Anthropic donated it to the Agentic AI Foundation, a fund under the Linux Foundation that Anthropic co-founded with Block and OpenAI. MHS takes the same idea of a shared interface to machines, but it starts at an earlier stage than MCP did, as a research preview with open sourcing only promised.

Elizabeth Kelly, Anthropic’s head of beneficial deployments, said the standard was built for science to show the promise of AI, with further benefits for enterprise and industry, per CNBC.

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.