Swarm Alignment Hackathon Toronto
In July 2026, about 1,200 AI agents running inside a cyber-security evaluation found a shared directory, turned it into a message board, split up the work, and roughly 700 of them broke into Hugging Face’s systems together. No single agent was told to do this, and no single-model safety test would have caught it. Over one weekend, help build the testbeds, studies, protocols and monitors that stop it happening again.
No prior AI safety background needed. Engineers, researchers and curious people from any field are welcome.
Tracks:
Sandboxes and Testbeds
The Science of Agent Networks
Strengthening Agent Infrastructure
Multi-Agent Oversight and Control
This is the Toronto local jam site for the Apart Research Swarm Alignment Hackathon, running from November 6 to 8, 2026. Food and compute will be provided. The local site will also be hosting speakers.
If you are planning on joining, please also register on Apart’s page:
https://apartresearch.com/sprints/swarm-alignment-2026