Discussion session: AIxBio

​Advances in AI and synthetic biology are rapidly changing the landscape of biosecurity risks, including the potential for engineered pandemics. In this session, we'll explore this in the context of what current AI systems can (or cannot) do and how their risks can be reduced.

​Our schedule:

• ​18:00–18:15 — Arrival and snacks

• ​18:15–18:45 — Presentations by:

  • ​Elias Nehring · Computational Biologist from ETH Zürich, ERA AI×Biosecurity Research Fellow
    Topic: Generative design of novel bacteriophages with genome language models

  • ​Samuel Šimko · Computer Scientist at ETH Zürich, incoming ELLIS PhD student
    Topic: Improving Large Language Model Safety with Contrastive Representation Learning + his recent ICLR 2026 work "Training with Honeypots: Reshaping How LLMs Fail"

• ​18:45 onward — Panel discussion and Q&A
  
  Having a look over the referenced papers is encouraged, but not expected! You're very welcome to also join spontaneously.

​​📍 How to join us
Zurich: PEAKS, Zeughausstrasse 31 — ring the "PEAKS" doorbell and wait for 30 seconds for someone to open the door.
Online: https://ethz.zoom.us/j/64473712209

For live coordination, join our Whatsapp group