FACEIT is rolling out a machine-learning system called “Human Input Detection” for its competitive Counter-Strike 2 platform, launching August 5, 2026, timed with the start of Season 9. Instead of scanning for known cheat software, the system analyzes in-game inputs to determine if they are authentically human. As of May 2026, AI and DMA cheats accounted for 40% of all bans on FACEIT’s platform. During Season 7, AI cheat detections spiked 272%, resulting in 372 bans. FACEIT’s existing kernel-level anti-cheat improved its coverage of AI-supported cheats through mid-2026, but kernel-level access alone was not enough. The new system is trained on millions of matches worth of input data, studying how humans move a mouse, click, and adjust aim to build a model of what “normal” looks like. FACEIT has been working with Google engineers on behavioral modeling since early 2026. The system requires multiple confirming signals before any ban is executed, and it includes accommodations for accessibility devices such as adaptive controllers and alternative input methods to avoid false positives.