The fine article bounces back and forth between base salary and total compensation (salary + bonus + stock grants). It’s hard to be certain of which they are actually using, but I work in Silicon Valley and this seems more like total comp than base pay. (Stock and bonuses can more than double total comp.)
Detectors of any sort can only flag expected variations from expected norms. AIs’ goals are to be undetectable with continuing improvements. Detectors help them do this by flagging failures. This is the same way antibiotic resistant bacteria evolve (well, it’s similar).