ITI SWITCH 2026 Seminar
November 20th 2026
Conseil room, EOST Bâtiment Descartes
5 rue René Descarte
From AI Prediction to One Health: Connecting Human and Environmental Toxicology for Sustainable Chemical Safety
Jinhee Choi
School of Environmental Engineering, University of Seoul, Seoul, South Korea
With a contribution from Donghyeon Kim,
PhD student in the dual-degree programme between the University of Seoul and Université Paris Cité,
who is closely involved in the development of the ToxBAI platform.
Abstract
Chemical safety assessment has traditionally been divided between human health and environmental protection, although chemicals released from consumer products, urban activities, and industrial processes move continuously across these boundaries through water, air, food, and biological systems. This seminar introduces ToxBAI, an adverse outcome pathway (AOP)-informed and mechanistically explainable AI platform designed to connect toxicity prediction with biological interpretation and decision-making.
ToxBAI integrates regulatory toxicity data, bioactivity information, and mechanistic knowledge to predict potential hazards and identify biologically plausible pathways. Using examples involving household biocides, PFAS, plastic-related chemicals, and other emerging contaminants, I will illustrate how the platform can support chemical prioritisation, the assessment of safer alternatives, and the translation of environmental toxicity evidence into potential human-health relevance. This approach also provides a basis for examining shared mechanisms of toxicity across species and for connecting ecotoxicology with human toxicology within a One Health framework.
The seminar will also address the limitations of AI-based assessment, including data gaps, model applicability, uncertainty, and the need for transparent validation. Finally, it will discuss how ToxBAI and related new approach methodologies can support interdisciplinary collaboration and more sustainable chemical management within water and urban systems, contributing to a safer and more sustainable society.
The seminar will also include a short practical presentation on ToxBAI, exploring how its models are developed, interpreted, and applied.
Discover ToxBAI, an AI-driven platform connecting toxicity prediction with biological mechanisms and chemical safety assessment. Through examples involving PFAS, biocides and emerging contaminants, explore how AI can bridge human and environmental toxicology within a One Health framework. A practical insight into the development, interpretation and application of ToxBAI will also be provided by Donghyeon Kim, PhD student at the University of Seoul and Université Paris Cité.
Keywords
Toxicology; One Health; AI-driven Chemical Safety
