Written by Satu Korhonen / April 4th 2026
There were many reasons I started to build games for people to play with breaking AI. One was to allow especially cyber security specialists to do things with AI that are usually not allowed in fair use principles. Another was to study how they do this and use this research to increase the understanding of how these models work and how we can build safer systems.
That choice has led to a fairly unique dataset of hacking attempts in several sizable events, like Disobey, Disarray, Easterhegg, and Balccon and so forth as well as use outside the events. The idea is to take this dataset and the framework for testing our games and create an open source version of them. Because, even as we are constantly kept busy doing interesting things, there is only so many hours in the day. By open sourcing these tools we can assist many others in creating safer AI than we ever could just by ourselves. This work is supported by FUUG and you can find more about that support on their blog here. So stay tuned. It’s going to be interesting.