My Ph.D. dissertation was on human and AI collaboration so I may have a unique perspective. Those interested may find comfort knowing that AI projects do not work if they do not clearly, and without question, work to advance humanity. So right now, with the general motorsports community having little interest in robots racing, our current state of motorsport will be safe for the next few years at least.
However, with continuing advancement in sensors and algorithms and the eventual application of quantum computing, it is reasonable to believe that aspects of autonomous vehicles will permeate into the most competitive of motorsports series (like F1) where AI could help the driver drive the car. I don't find that attractive at all but it'll be difficult to manage the technology if it makes the entire package faster. But for things like accurately predicting a tire deflating and ripping apart or when an engine is about to expire or minimize the damaging effects of a crash, etc., AI could serve a valuable purpose.