The rapid advancement of machine technology presents an novel and critical challenge: AI compromise. Cybercriminals are increasingly exploring methods to exploit AI platforms for illegal purposes. This involves everything from poisoning development data to bypassing security protections and even using AI-powered breaches themselves. The potential e