NEWS AND VIEWS
30 June 2026 AI systems devise hypotheses and ways to test them AI agents collaborate to generate biomedical hypotheses and analyse data, moving towards a laboratory discovery cycle with AI involved in every step. By Olivier Elemento ORCID: http://orcid.org/0000-0002-8061-9617 0 Olivier Elemento Olivier Elemento is in the Englander Institute for Precision Medicine, Department of Systems and Computational Biomedicine, Weill Cornell Medicine, New York City 10021, USA. View author publications PubMed Google Scholar
Scientific discovery has always been a deeply human enterprise, combining creativity, accumulated knowledge, teamwork and experimental ingenuity. Yet it is constrained by what scientists can collectively read, analyse and propose. Two papers in Nature, one by Gottweis et al.1 and one by Ghareeb et al.2, test how much further these capabilities can be pushed by multi-agent AI tools — systems consisting of several autonomous AI agents that cooperate to solve complex problems.
doi: https://doi.org/10.1038/d41586-026-01873-2
References Gottweis, J. et al. Nature https://doi.org/10.1038/s41586-026-10644-y (2026). Ghareeb, A. E. et al. Nature https://doi.org/10.1038/s41586-026-10652-y (2026). Boiko, D. A., MacKnight, R., Kline, B. & Gomes, G. Nature 624, 570–578 (2023). Swanson, K., Wu, W., Bulaong, N. L., Pak, J. E. & Zou, J. Nature 646, 716–723 (2025). He, L. et al. Cell 188, 6636–6653 (2025). Yanai, I. & Lercher, M. J. Nature Biotechnol. 43, 667–668 (2025). Cui, H. et al. Nature Methods 21, 1470–1480 (2024). Download references
Competing Interests O.E. is a co-founder of and stockholder in Volastra Therapeutics; holds stock in Freenome; serves on the Scientific Advisory Boards of and holds stock options in Owkin, Harmonic Discovery and Exai; and serves on the Scientific Advisory Board of Canary Biosciences. He has received research funding from Eli Lilly, Janssen/Johnson & Johnson, Sanofi, AstraZeneca and Volastra Therapeutics.
Related Articles
Subjects