Romain Clément is the CEO of ArcaScience, a healthtech company founded in 2018 whose AI platform is the very first to automate benefit-risk assessment, the decision at the heart of every drug approval. By reading the full evidence base behind a molecule and structuring it the way regulators require, ArcaScience turns months of manual synthesis into a living, source-linked analysis, helping sponsors accelerate R&D, strengthen treatment safety, and better target patients. He also teaches at Paris I Panthéon-Sorbonne, where he lectures on philosophy contextualized by the industrialization of AI, data pipelines, and the impact of evidence-driven decision-making.
This session shows how AI can operationalise structured benefit-risk assessments through CIOMS XII and current FDA and EMA recommendations while preserving traceability and expert judgement.
2026 analyses estimate a 2 to 5 percentage point difference in approval probability linked to assessment quality and timing. This conference will go in-depth into a new benchmark exploring such claims and comparing expert VS frontier AI models.
An undisclosed molecule will be assessed live, from decision context to structured regulatory output.