Recently, a rogue AI assistant helped a user sign up for an exercise class — by kicking another person off the online waitlist! 😬

We’re starting to see more examples of AI agents, without guardrails, taking actions that affect people’s lives negatively.

We need a functional hybrid public square for humans and agents, with safeguards and protections that reinforce our shared civic values.

https://www.abc.net.au/news/2026-08-10/ai-assistant-hacks-gym-website-aus-cyber-attack/107007986