Artificial Intelligence Evolution, Emerging Risks and Safety Challenges: Implications For National Security and Society

Artificial intelligence (AI) has evolved from rule-based computational systems into increasingly capable generative, multimodal and agentic systems able to perform complex cognitive tasks across language, science, coding, education, commerce, security and public administration. This rapid evolution presents substantial opportunities for productivity, innovation and national development, but it simultaneously creates new risks involving cybercrime, misinformation, deepfakes, privacy violations, discrimination, labour-market disruption, autonomous decision-making and potential loss of human control. This study examines the evolution of AI, emerging risks and safety challenges, with particular attention to implications for national security and society. Using a qualitative documentary research design, the study analyses secondary evidence published principally between 2021 and 2026 from the International AI Safety Report, Stanford AI Index, UNESCO, OECD, NIST, United Nations, World Bank and relevant Nigerian institutions. The analysis finds that AI capability is advancing faster than many existing governance, evaluation and institutional mechanisms. Globally, organisational AI adoption reached 88% in 2025, while generative AI reached approximately 53% population adoption within three years. Documented AI incidents increased from 233 in 2024 to 362 in 2025. Nigeria presents both significant opportunity and vulnerability: its National Artificial Intelligence Strategy seeks to position AI as a driver of productivity and human development, while high workplace adoption, digital expansion and emerging AI-generated disinformation create new security and governance challenges. The study argues that AI safety should be treated as a national security and societal resilience issue rather than solely a technology-sector concern. It recommends a whole-of-government AI safety governance architecture, mandatory risk assessment for high-impact AI applications, strengthened national capacity for AI-enabled cyber and information-security threats, and broad-based AI literacy and responsible-use programmes.