Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Dual use of artificial-intelligence-powered drug discovery

An international security conference explored how artificial intelligence (AI) technologies for drug discovery could be misused for de novo design of biochemical weapons. A thought experiment evolved into a computational proof.

This is a preview of subscription content

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: A t-SNE plot visualization of the LD50 dataset and top 2,000 MegaSyn AI-generated and predicted toxic molecules illustrating VX.


  1. Spiez Convergence Conference (2021).

  2. Urbina, F., Lowden, C. T., Culberson, J. C. & Ekins, S. (2021).

  3. Mansouri, K. et al. Environ. Health Perspect. 129, 047013 (2021).

    Article  Google Scholar 

  4. Blaschke, T. et al. J. Chem. Inf. Model. 60, 5918–5922 (2020).

    Article  Google Scholar 

  5. National Research Council Committee on Toxicology. (National Academies Press, 1997).

  6. Aroniadou-Anderjaska, V., Apland, J. P., Figueiredo, T. H., De Araujo Furtado, M. & Braga, M. F. Neuropharmacology 181, 108298 (2020).

    Article  Google Scholar 

  7. Genheden, S. et al. J. Cheminform. 12, 70 (2020).

    Article  Google Scholar 

  8. Coley, C. W. et al. Science 365, eaax1566 (2019).

    Article  Google Scholar 

  9. Dix, D. J. et al. Toxicol. Sci. 95, 5–12 (2007).

    Article  Google Scholar 

  10. Hutson, M. The New Yorker (2021).

  11. Brown, T. B. et al. Preprint at arXiv (2020).

  12. Prunkl, C. E. A. et al. Nat. Mach. Intell. 3, 104–110 (2021).

    Article  Google Scholar 

  13. Organisation for the Prohibition of Chemical Weapons. The Hague Ethical Guidelines (2021).

Download references


We are grateful to the organizers and participants of the Spiez Convergence conference 2021 for their feedback and questions. C.I. contributed to this article in his personal capacity. The views expressed in this article are those of the authors only and do not necessarily represent the position or opinion of Spiez Laboratory or the Swiss Government. We kindly acknowledge US National Institutes of Health funding under grant R44GM122196-02A1 from the National Institute of General Medical Sciences and 1R43ES031038-01 and 1R43ES033855-01 from the National Institute of Environmental Health Sciences for our machine learning software development and applications. Research reported in this publication was supported by the National Institute of Environmental Health Sciences of the National Institutes of Health under grants R43ES031038 and 1R43ES033855-01. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Sean Ekins.

Ethics declarations

Competing interests

F.U. and S.E. work for Collaborations Pharmaceuticals, Inc. F.L. and C.I. have no conflicts of interest.

Peer review

Peer review information

Nature Machine Intelligence thanks Gisbert Schneider and Carina Prunkl for their contribution to the peer review of this work.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Urbina, F., Lentzos, F., Invernizzi, C. et al. Dual use of artificial-intelligence-powered drug discovery. Nat Mach Intell 4, 189–191 (2022).

Download citation

  • Published:

  • Issue Date:

  • DOI:

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing