About Past Issues Editorial Board

KAIST
BREAKTHROUGHS

Research Webzine of the KAIST College of Engineering since 2014

Spring 2024 Vol. 22
Computing

AI for detecting aimbots in FPS games

BotScreen is a fully distributed AI system that can detect aimbots in FPS games in a scalable and reliable manner using a trusted execution environment. The research team won the Distinguished Paper Award at the USENIX Security Symposium 2023   Cheating in online games poses significant threats to the game industry.  Cheaters can directly impact the revenue of game publishers by annoying benign players and driving them away from the game. Cheating also undermines the fairness of the game, which can directly affect the Esports industry, which is now a billion-dollar industry. Therefore, it is imperative to design a systematic way to detect cheaters in online games.   First-person shooter (FPS) games are no exception. In particular, aimbot is one of the most significant cheating tools in FPS games. Aimbots enable a user to automatically aim at the opponent, which gives the user an unfair advantage over other players. Experienced players as well as novices can use aimbots to improve their performance in the game. Even professional players have been known to get caught using aimbots in tournaments.   There have been mainly two approaches to detecting aimbots:server-side and client-side approaches. However, both approaches have their limitations.  Server-side approaches are not scalable and require  many resources to process the data. Client-side approaches are not reliable because they can be easily bypassed by modifying the game client.   To address these challenges, Prof. Sang Kil Cha and his research team at KAIST, Cyber Security Research Center (CSRC) and Graduate School of Information Security (GSIS), proposed a fully distributed AI system that can detect aimbots in FPS games using a trusted execution environment (TEE). The proposed system, BotScreen, is fully distributed, which means that it does not require any centralized server to process the data, thereby making it scalable. The system overcomes the reliability issue of client-side approaches by using TEE, which is a hardware-based security technology that can protect the integrity of the code and data.   The key intuition of BotScreen's detection model is that aiming movements with or without aimbots are significantly different even though such differences are not visible to the human eye. Therefore, BotScreen builds an AI model by observing the aiming movements of benign players and then detects anomalies in the aiming movements of players to find cheaters.   The results of the study are published in a paper titled "BotScreen: Trust Everybody, but Cut the Aimbots Yourself" at the USENIX Security Symposium 2023. The paper was awarded the Distinguished Paper Award, which is given to the top 1% of all accepted papers. This work was supported by Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No.2021-0-01332, Developing Next-Generation Binary Decompiler).

Read more

Subscribe to our research webzine

Be the first to get the latest advancements in science and technology directly in your inbox.