CuspAI uses AI to design molecular sponges that remove PFAS forever chemicals from water supplies

CuspAI, led by machine learning researcher Max Welling, is using generative AI to design metal-organic frameworks (MOFs) that act as molecular sponges to trap and remove PFAS chemicals from water. PFAS, known as forever chemicals, do not break down naturally and are linked to serious health issues. The AI simulates millions of molecular structures to find ones that bind PFAS while letting clean water pass through. New EPA rules in the US now require municipal water systems to filter PFAS, creating a multibillion-dollar market for effective and affordable filtration. CuspAI's approach aims to be cheaper and more scalable than current methods like reverse osmosis. The technology could also help regions in the Global South, including East Africa, address industrial water contamination without building expensive traditional treatment plants.