Give game defenders the AI advantage.
We are working to bring frontier AI into game security, starting with tools to test game systems and evaluate their defences.
Meet the foundersIf attackers can remove the guardrails, how do defenders get ahead?
Open models can be adapted without a provider’s safeguards, while legitimate security work can face tighter constraints. We are working on practical tools and links with frontier labs so game developers can put specialist capabilities to use: investigating weaknesses, reproducing findings and testing stronger defences.
Our research direction