Skip to main content

Valve is teaching an AI to find Counter-Strike cheaters

valve counter strike cheater ai 561252 10150615540556283 2142228383 n cropped
Valve really doesn’t like cheaters, and the company is working on an artificial intelligence to combat them head on.

In a Reddit post, Valve confirmed that it has started work on an AI that is learning the difference between cheaters and highly skilled players. The AI has already started reporting bad eggs.

On the Global Offensive Subreddit, users posited why Valve had not created some kind of safeguard against cheaters who use spinbots. Valve actually chimed in to explain the difficulty in dealing with hackers.

“Any hard-coded detection of spin-botting leads to an arms race with cheat developers — if they can find the edges of the heuristic you’re using to detect the cheat, the problem comes back,” replied Valve_Anti-Cheat on Reddit.

A spinbot is a type of hack that spins a character so fast that it gives a 360 degree field of view. Because the player looks like they’re glitching, it also makes them a difficult target to hit. Couple this with aiming software and cheaters can essentially hit any player from any angle. It also makes spectating Counter-Strike pretty annoying too.

Unfortunately, catching these hackers is not so easy, as an AI has to evaluate each player. Multiply that by the over one-million matches that take place every day, and the system would quickly fall behind. Valve would essentially need to create a datacenter powering thousands of CPU cores just to catch cheaters.

Luckily, Valve does have an early version of a system already running and deployed. It’s monitoring games and submitting cases to Overwatch (not the game), Valve’s community group that looks at reports of disruptive behavior. If Valve’s new AI starts sending consistent and accurate reports, then Overwatch can start temporarily or permanently banning cheaters.

Unfortunately, until this AI is fully operational, cheaters will continue to ruin it for everybody else.

Editors' Recommendations