Artificial Intelligence (AI) is a transformative technology with significant potential and accompanying threats. While AI fosters innovation, it is increasingly exploited for criminal activities, especially cybercrimes. This study examines the criminal elements of AI, focusing on its role in enabling sophisticated cyber threats such as automated phishing, AI-driven malware, and deepfake fraud. Addressing these challenges requires organizations to adopt updated and actionable measures as traditional cybersecurity frameworks struggle to keep pace with rapidly evolving AI-driven threats.

This study identifies and categorizes the criminal elements of AI, with particular attention to their application in cybercrime. Requirements derived from expert interviews are the foundation for developing a conceptual framework to provide organizations with practical measures to combat AI-driven threats.

The results highlight the need for adaptable cybersecurity strategies that address both current and emerging challenges posed by offensive AI. The proposed framework, refined through expert feedback, offers targeted, actionable activities to enhance organizational resilience against AI-driven threats. By providing a clear structure and relevant measures, the framework supports cybersecurity professionals, IT managers, and risk officers in mitigating the dual-use risks of AI. This study contributes to bridging the gap in cybersecurity literature by addressing the evolving nature of AI-related threats and presenting a forward-looking framework tailored to middle-to-large organizations. It equips stakeholders with the tools necessary to safeguard digital ecosystems while ensuring the responsible use of AI technologies.