A*STAR and Bezos-Backed CuspAI Partner to Accelerate AI-Driven Materials Discovery in Singapore

Singapore's Agency for Science, Technology and Research (A*STAR) has entered a five-year partnership with CuspAI, a frontier AI materials science company backed by Jeff Bezos, to accelerate the discovery and experimental validation of new materials using artificial intelligence and autonomous laboratories. The collaboration spans semiconductors, carbon capture, and advanced electronics — placing Singapore at the centre of a growing global push to apply AI to physical sciences rather than just language and vision tasks.

CuspAI, founded in 2024 by chemist Chad Edwards and AI researcher Max Welling, describes its platform as a search engine for materials. Instead of testing candidate compounds one by one in a laboratory, its AI models predict which new materials should have desired properties, narrowing years of trial and error into a shortlist worth manufacturing. The company is raising approximately US$400 million from Bezos Expeditions and Kleiner Perkins at a roughly US$2.6 billion valuation, and has formed an AI Materials Foundry coalition that includes Nvidia, Meta, and Hyundai Motor Group.

The A*STAR collaboration is one of the first projects under CuspAI's Foundry programme and integrates Singapore's existing strengths in semiconductor research and advanced manufacturing. For Singapore, the partnership builds on the Research, Innovation and Enterprise (RIE) 2025 plan's investments in advanced manufacturing and materials science, adding an AI layer that could dramatically shorten the timeline from materials discovery to commercial deployment. This is particularly relevant as Singapore positions itself as a critical node in the global semiconductor supply chain, with wafer fabs from Globalfoundries and a growing cluster of chip design and materials firms.

The broader significance lies in what CuspAI represents: a shift from AI models that generate text or images to AI that accelerates physical discovery. For Asia, where foundries and battery manufacturers dominate global production, faster materials discovery directly benefits the region's core industries. However, the dependency risk is real — if the most powerful discovery software remains owned abroad, Asian manufacturers could find themselves paying for access to tools that shape their own supply chains.

Why it matters for Singapore: A*STAR's partnership with CuspAI signals that Singapore is not content to be a passive consumer of AI-driven materials discovery — it wants a seat at the table where the tools are built. The five-year commitment gives Singapore's research ecosystem direct access to frontier AI models for materials science, which could accelerate work at the Institute of Materials Research and Engineering (IMRE) and other A*STAR labs. For Singapore's semiconductor ambitions, the ability to rapidly validate new materials locally — rather than sending samples abroad — strengthens the case for keeping more of the chip value chain within the country.