CYBERSECURITY READINESS FOR ARTIFICIAL INTELLIGENCE: A COMPARATIVE ANALYSIS OF NATIONAL AI POLICIES AND CYBERSECURITY EDUCATION

The coverage of cybersecurity readiness for AI is uneven and under-theorised across countries, as they compete and develop their own national capacity-building strategies for AI. The paper provides a comparative policy review of AI policies and cybersecurity education programs in six jurisdictions, United States, EU, United Kingdom, Singapore, India, and Nigeria (in West Africa and African Union context). The review utilizes thematic synthesis to compare governance models, targets for the cybersecurity workforce pipeline and translation mechanisms for curricula, across thirty peer-reviewed and grey literature sources, such as national strategy documents, agency frameworks like ENISA, the NIST NICE Framework, and the Nigeria National Information Technology Development Agency (NITDA), as well as systematic reviews of curricula in cybersecurity. In every national AI strategy, cybersecurity is a national priority, but it is seldom defined in terms of competencies, and when it is, it is rarely addressed in terms of instruction, evaluation or certification. The centralised systems, like Singapore, are characterised by a stronger coupling of strategy to curriculum, while the federated or co-regulatory systems, like the United States and the European Union, exhibit weaker coupling, and Nigeria exemplifies a third model, where there are a highly ambitious curricular timeframe set in a central manner, but significant capacity constraints to support the workforce and a less unified regional legal framework. There is limited capacity for assessing educational interventions in each of the jurisdictions studied. The paper recommends a crosswalk between national AI policies and cybersecurity education competency frameworks, as well as priority areas for future research and policy design, such as cross-border harmonisation, including at the regional level in Africa, in particular the African Union, equity-oriented implementation, and longitudinal outcome evaluations.