Frontier stage
Keynote
In person

Demo: AI deception amplifies other risks

  • Date
    10 July 2025
    Timeframe
    14:30 - 15:00 CEST
    Duration
    30 minutes

    Drawing on our experience conducting pre-public deployment testing of frontier AI systems, Apollo Research offers insights into the risks posed by deceptive AI. We will do a live showcase of several evaluations we’ve conducted to detect deception, along with their results.

    Are you sure you want to remove this speaker?