UST is bringing Claude to physical AI

UST is integrating the Claude AI model into its engineering and production environments to advance physical AI, which involves embedding intelligence into the manufacturing and design processes for chips, cars, and connected devices. By incorporating Claude into its existing platforms, UST aims to catch design flaws earlier in the development cycle, reducing the high costs associated with errors discovered during mass production.

A primary application of this technology is the iDEC platform, which validates hardware and silicon. Claude acts as a reasoning layer that reads schematics and pinouts to write and run regression tests automatically. This process has already reduced validation cycle times by 50 to 70 percent. Beyond hardware, UST is deploying Claude into its healthcare, telecom, and banking platforms to streamline claims management, predict network outages, and modernize legacy banking systems through automated case handling and workflow support.

To support this initiative, UST is training 20,000 of its engineers, architects, and consultants worldwide on Claude. The partnership emphasizes safety and governance, ensuring that human oversight remains a core component of all AI-driven decisions. By embedding Claude directly into these high-stakes industrial workflows, UST intends to improve productivity and accelerate business outcomes for its clients while maintaining the rigorous security standards required in regulated industries.