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.


Why deep-learning AIs are so easy to fool

Illustration by Edgar Bąk

Access options

Rent or buy this article

Prices vary by article type



Prices may be subject to local taxes which are calculated during checkout

Nature 574, 163-166 (2019)



  1. Eykholt, K. et al. IEEE/CVF Conf. Comp. Vision Pattern Recog. 2018, 1625–1634 (2018).

    Article  Google Scholar 

  2. Finlayson, S. G. et al. Science 363, 1287–1289 (2019).

    Article  PubMed  Google Scholar 

  3. Elsayed, G. F., Goodfellow, I. & Sohl-Dickstein, J. Preprint at (2018).

  4. Szegedy, C. et al. Preprint at (2013).

  5. Nguyen, A., Yosinski, J. & Clune, J. IEEE Conf. Comp. Vision Pattern Recog. 2015, 427–436 (2015).

    Article  Google Scholar 

  6. Alcorn, M. A. et al. IEEE Conf. Comp. Vision Pattern Recog. 2019, 4845–4854 (2019).

    Google Scholar 

  7. Hendrycks, D., Zhao, K., Basart, S., Steinhardt, J. & Song, D. Preprint at (2019).

  8. Huang, S., Papernot, N., Goodfellow, I., Duan, Y. & Abbeel, P. Preprint at (2017).

  9. Gleave, A. et al. Preprint at (2019).

  10. Silver, D. et al. Science 362, 1140–1144 (2018).

    Article  PubMed  Google Scholar 

Download references


Latest on:

Nature Careers


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


Quick links