Online competitive games like Counter-Strike: Global Offensive are rife with cheating players, but finding a way to catch those players in the act is no simple feat, but well trained and adaptable AI could offer game developers a solution to the problem.
Current systems, which rely on individually reviewed player reports, are labor intensive and only take down cheaters after the fact while any “hard-coded” cheat detection leads to what one Valve employee described as an arms race between cheaters and developers.
Valve’s solution? Create and train a learning AI to detect differences in gameplay between legitimate players and cheaters.
This information comes way of a Reddit comment from Valve originally picked up by Kotaku, who said that early versions of a cheat-detecting AI have already shown promise but that the process is both complicated and very resource intensive.
“The process of parsing, training, and classifying player data places serious demands on hardware, which means you want a machine other than the server doing the work,” explained Valve. “And because you don’t know ahead of time who might be using this kind of cheat, you’d have to monitor matches as they take place, from all ten players’ perspectives.”
“There are over a million CS:GO matches played every day, so to avoid falling behind you’d need a system capable of parsing and processing every demo of every match from every player’s perspective, which currently means you’d need a datacenter capable of powering thousands of CPU cores.”
And while this system is understandably demanding, Valve may be one of the companies best equipped to devote resources to such a project. The company says it’s already started work on developing this technology. A very early version of it was quietly dropped into Counter-Strike: Global Offensive at some point and the agent has already had some success submitting its own reports of cheating players.