Salman Shaikh
Ph.D. Student
King Abdullah University of Science and Technology (KAUST)
Salman Shaikh is a Ph.D. student in Computer Science at KAUST (King Abdullah University of Science and Technology), where his primary focus is on understanding and researching security in Multiplayer Online Games. His research sits at the intersection of game development, networking, attack exploitation, and defensive strategies, with the goal of providing detection and mitigation techniques for today's unsolvable cheats that exploit the trust model the game industry relies upon.
Speaker sessions
Briefings and Arsenals | Confirmed Speakers | No Session Assigned
Placeholder session for confirmed speakers without an assigned session, allowing them to be pushed to Swapcard and Visit.
- 07:30
- Tue
- 01 Dec
Stage:
Briefing Stage 2
Sessions Type:
Keynote
Visual Aimbots: How they evade current anti-cheat systems, and how can we catch it?
Aimbots have plagued multiplayer online games for a very long time. Traditionally, they operate by reading the game's memory, by extracting critical data structures, performing necessary calculations and injecting input into the memory. However, they are typically battled via current commercial anti-cheat systems that operate at the kernel level and detect any kind of memory tampering. But the latest generation of AI object-detection models opens an entirely new threat surface, one that lives outside the client altogether and completely invisible to anti-cheat systems. In this session we show the reality, and the severity of this threat. We built a fully working visual aimbot from scratch, starting from its training to its real-time operation. At run-time, instead of exploiting the game memory, it simply looks at the rendered screen, feeds frames into an object detector, locates enemy players automatically and snaps the crosshair onto them. The consequence is extreme. This attack is undetectable to any currently proposed defense in the industry and in the academic community. This is primarily because there is nothing for systems to detect or any unpredictable behavior to analyze. Confronted with this, we propose a solution that flips the cheaters' own arsenal against them. We introduce PATCH, which are adversarial honey-tokens that remain as invisible noise to honest players but dishonest players who use visual aimbots, their model cannot resist targeting these patches, baiting the cheaters into revealing themselves. We close with results, showing the efficacy of our patches, and its applicability on a real commercial game.
- 13:50
- Thu
- 03 Dec
Stage:
Briefing Stage 2
Sessions Type:
Keynote