IBM Research has been exploring the application of AI to tackle pressing issues posed by materials of concern. By leveraging advanced foundation models for scientific domains and refined agentic workflows, IBM has significantly accelerated corporations' ability to identify and mitigate these materials promptly. Across most many industries, businesses face prolonged manual identification of issues, followed by the challenging task of avoiding regrettable substitutions, and developing new alternatives.
One key category of materials of concern for companies contain per- and polyfluoroalkyl substances (PFAS) – synthetic “forever chemicals” that can be found in everything from nonstick pans, clothing, and food packaging, to furniture, electronics, and concrete. Nearly every industry has significant exposure to PFAS chemicals, which pose serious environmental and health concerns due to their persistence, toxicity, and bio-accumulative properties.
Ever-evolving regulations, inconsistent definitions and data availability issues make addressing PFAS a particularly challenging problem. New regulations targeting PFAS seem to appear almost daily, led by the European Chemicals Agency (ECHA) and the US Environmental Protection Agency (EPA). Many other countries have already implemented – or are proposing to implement – PFAS restrictions. Organizations also face massive liability risk with PFAS: in the United States alone, over $13B in damages have been awarded to settle lawsuits. The European Union is also imposing hefty fines for environmental crimes.
While PFAS is just one of many materials of concern posing a threat to consumers and corporations alike, consumer awareness around hazardous materials in everyday products is growing, influencing purchase decisions and increasing pressure on companies to find safer alternatives. By leveraging new AI capabilities, IBM is accelerating the ability to assess and prioritize chemical risks and mitigation strategies, including the design of PFAS replacements and capture materials.
IBM Safer Materials Advisor is an easy-to-use AI application designed to identify materials of concern within products, manufacturing processes, and supply chains. Initially targeting PFAS, the advisor effectively detects PFAS compounds in various parts, components or formulations, offering a thorough evaluation of the most hazardous chemicals.
In future releases, the advisor will assist in sourcing substitutes — considering factors like performance, sustainability, and operational dependencies — and offer unique approaches based on transfer-learning of successful solutions from diverse industrial applications. In scenarios where no direct substitute is apparent, the AI-powered design capabilities of the advisor will facilitate the discovery of entirely novel materials, ensuring continuous advancement in the sourcing of safer materials.
The IBM Safer Materials Advisor offers:
This comprehensive approach ensures that organizations can effectively identify, assess, and mitigate the risks associated with PFAS, paving the way for safer and more sustainable material management strategies.
Creating the AI-enabled lab for a new era of reproducible and collaborative experimentation