Taxonomy of Cybersecurity Challenges and Solutions for the Superintelligent Systems

Context Contemporary developments in the domain of Artificial Intelligence (AI) have made the possibility of achieving superintelligence systems only a matter of time. This possibility presents both enormous opportunities and challenges in cybersecurity. This research present detailed taxonomy of the cybersecurity challenges posed by superintelligence. Objectives This research systematically synthesizes and classifies the theoretical cybersecurity challenges posed by the development of superintelligent systems and presents conceptual mitigation strategies drawn from formal and grey literature. Methodology This research is conducted using a Multivocal Literature Review (MLR). The search strategy was developed using the PICO (Population, Intervention, Comparison, Outcome) technique. To ensure reliability we utilized a five-phase tollgate approach, including quality assessment scoring for both formal and informal literature. A total of 82 research articles were reviewed, 49 of which were formal literature sources, and the remaining 33 were grey literature sources. Results This research has identified 13 major cybersecurity challenges and a total of 64 practices for the identified cybersecurity challenges posed by superintelligent systems. The results of this research will contribute to the ongoing developments in highlighting the security threats posed by advanced AI systems. Conclusion This research presents detailed accounts of the taxonomy of cybersecurity challenges and solutions in superintelligent systems.

Department

Electrical and Computer Engineering

Publisher

Elsevier

Sponsor

Effat University