The fast advancement of AI technology presents the new and serious challenge: AI compromise. Cybercriminals are ever more developing methods to abuse AI systems for illegal purposes. This includes everything from corrupting development data to circumventing security protections and even deploying AI-powered breaches themselves. The potential effect