Selected Publications

Juijn, D., van Baarsen, S., Dada, J., Negele, M., Stelling, L., Fox, P., Petropoulos, A., Bakker, M. (2026). “Europe 2031: What getting AI wrong means for us”, web version.

Bengio, Y., Clare, S., Prunkl, C., Murray, M., Andriushchenko, M., Bucknall, B., Bommasani, R., Casper, S., Davidson, T., Douglas, R., Duvenaud, D. Fox, P. [+ 81 authors], “International AI Safety Report 2026”, DSIT 2026/001.

Fox, P., Schnitzer, M., Privitera, D. (2025). “AI Compute in Germany: Current capacity, future needs”, KIRA Report, English version here, German version here.

Bengio, Y., Mindermann, S., Privitera, D., Besiroglu, T., Bommasani, R., Casper, S., Choi, Y., Fox, P., Garfinkel, B., Goldfarb, D. [+ 86 authors] (2025). “International Scientific Report on the Safety of Advanced AI”, arXiv:2501.17805 / DSIT 2025/001

Fox, P., Barrett, S., Krook, J., Mondal, T., Mylius, S., Tlaie, Al. (2025). “Assessing confidence in frontier AI safety cases”, arXiv:2502.05791

Other Publications

“The economics of frontier model access: a reply”, Substack, 23.06.2026 [this piece replies to a commentary by Dezernat Zukunft]

“5 Takes on Mythos”, Substack, 17.04.2026

“Electricity Prices: An Overrated Barrier to European AI Compute”, Substack, 08.04.2026

“Should Middle Powers Worry about Kill Switch Scenarios for Data Centers?”, Substack, 12.02.2026

“Making AI Progress More Legible”, Substack, 11.01.2026

“7 Learnings about Compute Strategy for Middle Powers”, Substack, 17.12.2025

Bengio, Y., Clare, S., Prunkl, C., Andriushchenko, M., Bucknall, B., Fox, P., Maslej, N., McGlynn, C. [+ 61 authors], “International AI Safety Report 2025: Second Key Update: Capabilities and Risk Implications”, arXiv:2511.19863, DSIT 2025/042.

“The EU’s Apply AI Strategy: Sensible, but not a breakthrough”, Table.Briefings, 27.10.2025

Bengio, Y., Clare, S., Prunkl, C., Rismani, S., Andriushchenko, M., Bucknall, B., Fox, P., Hu, T., Jones, C. [+ 64 authors], “International AI Safety Report 2025: First Key Update: Capabilities and Risk Implications”, arXiv:2510.13653, DSIT 2025/033.

“Making the Case: How Can We Know When AI is Safe?” (with Ketana Krishna, Tuneer Mondal, Ben Smith & Alejandro Tlaie), AI Policy Bulletin, 06.08.2025

“How many Gigafactories will the EU build – and why?”, Substack, 28.07.2025

“There is No Risk of Misalignment”, Substack, 23.05.2025

“Red Lines in the Sand: Responsible Scaling Policies Need Quantitative Thresholds”, Substack, 10.04.2025

“When AI Policy is Handcuffed, Spend on Alignment”, Tech Policy Press, 18.03.2025

“Why Germany needs an AI Security Institute” (with Anton Leicht), Tagesspiegel Background, 17.03.2025

“An AI-risk expert thinks governments should act to combat disinformation”, The Economist, 06.02.2024

“Why AI is so hard to control”, Spektrum der Wissenschaft, 15.12.2023

“Foundation model regulation in the AI Act: A right step, but with gaps”, table.media, 12.12.2023

“How AI is threatening public discourse – and what to do about it” (with Anka Reuel), Frankfurter Allgemeine Zeitung, 17.09.2023

“How do Germans think about AI?” (with Daniel Privitera and Anka Reuel), KIRA Report, April 2023