The rapid advancement of machine technology presents a novel and serious challenge: AI compromise. Cybercriminals are ever more exploring methods to manipulate AI systems for harmful purposes. This encompasses everything from poisoning development data to bypassing security protections and even using AI-powered attacks themselves. The potential eff