Water/Wastewater
Kemira and CuspAI have announced a breakthrough in the fight against PFAS contamination, unveiling new AI-designed materials capable of removing “forever chemicals” from drinking water and industrial process water at trace concentrations.
The collaboration marks the first commercial partnership to use generative artificial intelligence end-to-end for designing entirely new materials specifically aimed at PFAS remediation. The project compressed a discovery process that traditionally takes years into just six months.
Using advanced generative AI models, the project explored an estimated 300 trillion possible material structures to identify high-performance candidates for removing harmful PFAS compounds from water.
The initiative produced more than 5,000 novel material designs with full property data targeting three priority PFAS molecules — GenX, PFBS, and PFOS. From these, approximately 20 high-priority candidates have now been selected for the next phase of development and testing.
The materials were designed to meet strict industrial requirements, including water stability, environmental compatibility, manufacturability, cost-effectiveness, and high removal efficiency at sub-parts-per-billion concentrations.
CuspAI’s platform generated entirely new metal-organic frameworks (MOFs) from scratch, creating advanced adsorption materials tailored specifically for PFAS capture.
PFAS, or per- and polyfluoroalkyl substances, are persistent synthetic chemicals linked to growing environmental and health concerns worldwide. Often referred to as “forever chemicals”, PFAS are difficult to remove from water supplies due to their chemical stability.
Current leading PFAS remediation technologies rely heavily on granular activated carbon (GAC). Kemira initiated the collaboration to investigate whether AI-driven materials design could produce more selective, durable, and longer-lasting alternatives.
The breakthrough comes as governments worldwide tighten regulations surrounding PFAS contamination.
In 2024, the US Environmental Protection Agency (EPA) introduced new maximum contaminant limits for PFAS measured at parts-per-trillion levels, while the European Union continues implementing stricter standards under the EU Drinking Water Directive.
As regulatory pressure grows, demand for scalable and sustainable PFAS removal technologies is expected to rise significantly across municipal water treatment and industrial sectors.
Dr Chad Edwards, CEO and Co-Founder of CuspAI, said the project demonstrates how AI can dramatically accelerate materials discovery for critical environmental challenges
According to Edwards, the partnership delivered meaningful results in just six months and has already moved into the next phase of development, with additional projects now being scoped.
Sampo Lahtinen, Executive Vice President of Research and Innovation at Kemira, emphasised that all material candidates were evaluated against real-world industrial performance requirements, helping ensure commercial viability.
Antti Salminen, President and CEO of Kemira, said the collaboration established a credible path towards next-generation PFAS remediation products capable of removing forever chemicals at extremely low concentrations while remaining sustainable and manufacturable at scale.
The project has now advanced into further testing and development, with researchers also identifying new functional group chemistries that may support broader adsorption technologies in the future.
Under the partnership’s framework agreement, Kemira and CuspAI are expected to explore additional material classes and continue expanding AI-driven innovation for water treatment applications.
The breakthrough highlights the growing role of artificial intelligence in accelerating clean technology innovation and addressing global environmental challenges such as PFAS contamination in drinking water.
Product directory
