Ahead of Rethinking Materials 2026, we sat down with Kristin Schmidt, Strategy Assistant, Algorithms and Applications, and Safer Materials Advisor Lead at IBM Research, to explore how artificial intelligence is helping companies identify chemicals of concern, navigate regulatory complexity, and move faster towards safer, more sustainable material choices.

In this video interview, Kristin discusses the challenges of managing PFAS and other high‑risk substances across global supply chains, why data gaps remain such a major barrier, and how AI can augment — not replace — human expertise in materials decision‑making.

👉 Watch the full interview below, or read on for the key questions and insights from the conversation.

Kristin, can you briefly introduce your role at IBM and how it connects to safer materials and sustainability?

Kristin Schmidt: I’m sitting at a very interesting intersection between sustainability and AI at IBM Research. I lead the Safer Materials Advisor, which is a tool that helps companies identify materials of concern in their products, assess both regulatory risk and chemical hazard, and then move towards replacing those substances with safer alternatives.

AI allows us to make this process much easier to use and, importantly, to scale it. Traditionally, this kind of work has been very manual and resource‑intensive, which makes it difficult for organisations to get a comprehensive view of chemical risk across their portfolios.

Kristin Schmidt: One of the biggest problems is actually the quality of information companies receive from their suppliers. Data is often inconsistent, incomplete, or obscured — in some cases compounds are omitted or hidden behind trade secrets or proprietary formulations.

PFAS makes this even harder because regulations differ. Some are definition‑based, others are list‑based, which means it’s not always clear what counts as a PFAS.

What we do is use AI to analyse documentation at scale. The system flags inconsistencies and missing information, and then suggests what those missing components might be — or at least which chemical class they fall into. That enables much more robust regulatory analysis and better assessment of chemical hazards, giving companies clearer insight into what’s actually in their products.

Kristin Schmidt: Once the AI has flagged gaps or risks, it looks for analogous products in public databases, patent databases, and other available material data sources. It compares ingredient lists and functional properties between products to infer what compounds might be present, or again, which chemical class they belong to.

It’s not about making the decision for the user - it’s about narrowing down the options and providing better, evidence‑based inputs so that companies can explore safer alternatives more efficiently.

Kristin Schmidt: It definitely augments human expertise - we don’t want to replace experts at all.

Before this tool, we saw sustainability teams spending huge amounts of time manually reviewing material safety data sheets, looking up compounds across multiple tools and websites, and documenting everything in spreadsheets. Now, the AI analyses documents at scale, connects directly to databases, and surfaces all the relevant information in one place.

Crucially, it always shows where the information comes from. That transparency builds trust, because users understand how conclusions are reached and can make informed decisions themselves.

Kristin Schmidt: At the end of the day, the responsibility still sits with the user — they make the final call. But they’re doing so with much better information.

What we also see is that the tool flags errors or blind spots that users wouldn’t necessarily have discovered on their own. That leads to a much more comprehensive understanding of chemical risk and regulatory exposure across an organisation.

Kristin Schmidt: There’s a real need for better collaboration across industries. Material data today is spread across different databases, documents, and formats, and it’s rarely standardised or harmonised.

Another major challenge is enabling suppliers to share critical safety information without exposing their intellectual property. AI has the potential to help here — by enabling the sharing of what matters for safety and compliance without revealing confidential formulations.

Initiatives like shared bills of materials and digital material passports are steps in the right direction, and we’re starting to see momentum build.

Kristin Schmidt: The diversity of people attending. Bringing together voices from technology, regulation, materials science, and sustainability creates opportunities to exchange ideas and identify where collaboration can really accelerate progress.