Machine learning (ML) is reshaping how environmental chemicals are monitored and how their hazards are evaluated for human health. Here, we mapped this landscape by analyzing 3150 peer-reviewed articles (1985–2025) from the Web of Science Core Collection. Co-citation, co-occurrence, and temporal trend analyses in VOSviewer and R reveal an exponential publication surge from 2015, dominated by environmental science journals, with China and the United States leading in output. Eight thematic clusters emerged, centered on ML model development, water quality prediction, quantitative structure–activity applications, and per-/polyfluoroalkyl substances, with XGBoost and random forests as the most cited algorithms. A distinct risk assessment cluster indicates migration of these tools toward dose–response and regulatory applications, yet keyword frequencies show a 4:1 bias toward environmental endpoints over human health endpoints. Emerging topics include climate change, microplastics, and digital soil mapping, while lignin, arsenic, and phthalates appear as fast-growing but understudied chemicals. Our findings expose gaps in chemical coverage and health integration. We recommend expanding the substance portfolio, systematically coupling ML outputs with human health data, adopting explainable artificial intelligence workflows, and fostering international collaboration to translate ML advances into actionable chemical risk assessments.
Stanić et al. (2025) studied this question.