MOZA Racing is putting its Racing Lab AI Coach in the spotlight at Gamescom 2026, using connected wheelbases and pedals to physically teach steering, braking and throttle techniques. An AI vs Human challenge featuring Jarno Opmeer and Dirk Schouten is scheduled for August 29.
MOZA Racing is using Gamescom 2026 to showcase the latest development of its Racing Lab AI Coach, an AI-powered training system that goes beyond conventional telemetry analysis by directly interacting with compatible sim racing hardware. Rather than simply showing where a driver is losing time, the system can control the steering wheel and pedals so users can physically experience the inputs generated by MOZA’s AI driver.
The technology itself is not being revealed for the first time at Gamescom. MOZA originally introduced Racing Lab at GDC 2026 on March 9, before providing further details during its July product announcements. Gamescom represents its latest major public demonstration, accompanied by an AI-versus-human event designed to show how its internally developed driver compares with experienced real and esports racers.
How MOZA AI Coach Works
Racing Lab uses an AI Driver developed in-house by MOZA and trained through reinforcement learning. According to the manufacturer, its reference driver has been developed to produce lap times comparable to, and in some cases faster than, professional-level drivers.
The main difference from traditional coaching applications is the connection between that AI model and MOZA hardware. Racing Lab can actively operate a compatible MOZA direct drive wheelbase and pedals, demonstrating steering angles, braking inputs and throttle application. A driver can lightly hold the controls and physically feel what the AI is doing, with the aim of building muscle memory rather than relying exclusively on graphs, videos or racing-line overlays.
Training can also be split between individual inputs. Users can practise steering while the AI manages the pedals, or work specifically on braking and throttle while receiving steering assistance. MOZA has outlined Beginner, Intermediate and Professional training levels, alongside different degrees of AI intervention as drivers move towards fully independent laps.
Three steering guidance methods have also been detailed: Active Guidance, Pulse Guidance and Haptic Guidance. These provide different levels of physical assistance depending on whether the driver wants stronger intervention or subtler feedback from the wheel.
Racing Lab additionally includes a real-time overlay comparing the driver’s inputs with the AI reference. Steering, throttle, braking, gear selection and racing line data can be displayed while driving, while recorded sessions can subsequently be compared through telemetry graphs to identify where the driver’s lap differs from the AI reference.
Importantly, MOZA states that AI Driving is intended for personal practice and demonstrations rather than racing on a user’s behalf. The active AI Driving function is disabled when Racing Lab detects a multiplayer session.
Assetto Corsa and ACC Are Currently the Main Focus
MOZA’s current Racing Lab documentation lists Assetto Corsa and Assetto Corsa Competizione as the two supported Racing Lab titles, with AI Driving, real-time comparisons and telemetry recording available across supported combinations.
AI reference data is generated individually for particular cars and circuits, meaning support is not automatically available for every combination within each simulator. MOZA currently lists ten tracks and ten vehicles for each game as its initial supported selection, including circuits such as Spa-Francorchamps, Monza, Silverstone, Nürburgring GP, Barcelona, Imola, Laguna Seca and Red Bull Ring.
Hardware requirements are similarly specific. Racing Lab requires a MOZA wheelbase and at least two MOZA pedals assigned to throttle and brake, while third-party pedals are not currently supported for AI Driving. MOZA’s mBooster active pedals can be directly controlled by the system, while supported mechanical MOZA pedals need to be connected through the wheelbase for the required functionality.
MOZA has been rolling Racing Lab out through Pit House, although its current dedicated Racing Lab page still provides an early-access registration process. The company has therefore not presented Gamescom as a separate paid AI Coach product with its own standalone price.
Jarno Opmeer and Dirk Schouten Will Take on the AI
The major new Gamescom element is an AI vs Human Driver Challenge scheduled for August 29 at MOZA’s booth in Cologne.
Three-time F1 Esports World Champion Jarno Opmeer and MOZA global ambassador and Porsche racing driver Dirk Schouten are set to participate in the demonstration, with MOZA planning to broadcast the session through its YouTube channel. The company has already confirmed its Gamescom activities, which take place at Hall 10.1, Booth A-022.
The challenge will provide a more visible test of Racing Lab’s AI driver and the broader concept behind the system: using AI not simply to analyse a completed lap, but to generate a reference that can be transferred back through physical controls and used as a training tool.
MOZA has not announced a separate commercial price for Racing Lab at Gamescom, nor has it confirmed expansion to additional simulators as part of today’s showcase. Its current documented support remains centred on Assetto Corsa and Assetto Corsa Competizione, with further combinations dependent on additional AI reference data being developed.
The AI vs Human Driver Challenge takes place on August 29, while MOZA’s Gamescom 2026 exhibition runs from August 26 to 30 in Cologne.
See you on the track!
This website uses affiliate links which may earn a commission at no additional cost to you.
