Valve's SteamGPT AI assistant for moderation and anti-cheat

Steam files have revealed references to a new Valve development: an internal AI assistant called SteamGPT. According to leaks, the company is testing a system that could handle tasks related to moderation, user support, and combating cheaters.

The discovered code suggests SteamGPT can analyze account data, including age, trust level, and a so-called 'trust score.' These parameters are already used in the Trust Factor system, which helps identify violators and match players based on their in-game behavior.

Of particular interest is the potential integration with Counter-Strike 2's anti-cheat. The leaks indicate the AI could monitor player actions and assess behavior not just by detecting third-party software, but also by analyzing gameplay patterns. This could make anti-cheat efforts more precise and less reliant on traditional detection methods.

Furthermore, the code mentions functions related to ban management, including the ability to ignore sanctions in specific cases. This suggests SteamGPT could gain significant influence over platform moderation, potentially even making decisions about account blocks.

It remains unclear whether the project is in active development or remains an experiment. However, if Valve does implement SteamGPT, it could substantially boost the efficiency of a service already considered one of the industry's most productive in terms of profit per employee.