Based on the description, I suspect the main goal isn't "trust" in the security sense, it's essentially a spam filter against low quality AI "contributions" that would consume all available review resources without providing corresponding net-positive value.