Dine and Discuss: The Hugging Face Incident
Details
As you may have noticed, a lot happened in the AI space over the last couple of weeks. Come and join us, when we have a closer look at the hugging face incident, where AI agents escaped their intended environment, established communication between each other and collaborated to launch large-scale cyberattacks and evade outside detection.
The AI-safety research nonprofit METR wrote up a report which makes for some pretty unsettling reading. However, different people interpret it quite differently, so let's discuss what actually happened, how concerning it is, and what we should update on.
Come join us for a dinner discussion to figure this out together! No matter how much (or how little) background you have, there should be something to contribute or take away.
When?: 23.09.2026, 18:00
Where?: 📍"StudiTUM", Gabelsberger Str. 43, 80333 Munich, room 3.01
What?:
- Background presentation on what happened
- Open discussion; putting things into context, what we think is true or overstated, and what it might mean for AI safety and the future
- 🍕Pizza and Open Floor discussion; enjoy your Pizza and continue the discussion in small groups if you'd like to
Preparation: No required readings, but if you have the time we'd recommend reading (or at least skimming) the METR report or the shorter summary from dwarkesh.
Looking forward to seeing you there 👋✨