CuspAI and AI: Revolutionizing Material Discovery

CuspAI: Transforming Material Discovery with AI

Chad Edwards and Max Welling, the brains behind CuspAI, aim to revolutionize the way materials are discovered and developed. Artificial intelligence, which has already disrupted the fields of coding and writing, is now extending to material manufacturing. In this sector, innovation is often a lengthy and uncertain process, but CuspAI intends to change this dynamic.

Rethinking the Laboratory Process

Traditionally, the creation of new materials relies on an empirical approach. Researchers formulate hypotheses, test them, and adjust their methods, sometimes over several years, without any guarantee of success. CuspAI proposes to reverse this logic by first focusing on the desired properties, such as conductivity, thermal resistance, and energy capacity. Based on these criteria, artificial intelligence generates optimal molecular configurations. This approach is part of the "AI for science" movement, which aims to integrate learning models at the heart of the scientific process. However, the promise of this method hinges on the ability to align simulation and experimental validation, two steps with distinct timelines and constraints.

Strategic Industrial Partnerships

CuspAI targets industries where materials play a crucial role in performance, such as semiconductors, batteries, energy, and carbon capture. In these sectors, even a slight advantage in physical properties can provide a significant competitive edge. The startup has already signed commercial contracts worth several tens of millions of dollars with renowned companies such as NVIDIA, ASML, and Hyundai Motor Company. These partnerships serve as real-world tests to evaluate the integration of AI into demanding industrial chains.

A Funding Momentum in Line with AI for Science

After an initial funding round in 2024, CuspAI raised a Series A in 2025 with investors like New Enterprise Associates and Temasek, reaching a valuation close to $800 million. Currently, a new fundraising round of at least $200 million is underway, aiming to achieve unicorn status. This momentum is part of a broader trend where AI for science is attracting increasing capital, particularly in the United States. Startups founded by former researchers from OpenAI and Google DeepMind have already reached valuations exceeding $1 billion, backed by influential investors like Jeff Bezos. However, European players are struggling to keep up due to a lack of capital and a less developed ecosystem.

Support from Leading Figures in Deep Learning

CuspAI benefits from top-tier support, both scientific and industrial. Among its advisors are major figures in artificial intelligence, such as Geoffrey Hinton and Yann LeCun, as well as Martin van den Brink, former president and CTO of ASML, and Lord John Browne, former CEO of BP.

The Challenge of Transitioning from Computation to Reality

The main challenge for CuspAI lies in the fact that the rapid generation of promising molecular structures does not replace laboratory testing cycles, certification processes, or industrial integration constraints. While computation can accelerate certain steps, it does not necessarily reduce the overall duration of the innovation cycle. Clients, particularly large industrial groups, operate on long horizons with high reliability requirements. The adoption of this tool will depend more on the accumulation of validated results than on mere promises. Furthermore, CuspAI's business model is still under construction, with options such as software licenses, SaaS platforms, and industrial co-development. No standard has yet emerged in this burgeoning sector.

An Indicator of a Broader Transformation

CuspAI represents a new generation of startups where artificial intelligence plays an increasingly central role, shifting from optimizing existing processes to intervening upstream in the very definition of industrial objects. This trend deserves the attention not only of industrialists and investors but also of public decision-makers, in order to address the challenges it raises. Beyond the case of a single startup, it reveals a strategic capability to design the materials that will shape the next generations of energy, digital, and industrial infrastructures, thus representing a direct lever of sovereignty.