This study explores how AI technologies enhance public scientific literacy through targeted education initiatives and interactive communication tools. Using a mixed-methods approach (2023–2025), it analyzes 11 AI-driven science education programs and 8 interactive tools across 9 urban centers (e.g., London, Delhi, Cape Town). Results show AI-assisted education increases scientific knowledge retention by 52% vs. traditional methods, with interactive tools (e.g., AI chatbots, VR science labs) boosting engagement by 47% among low-literacy groups. Key predictors of success include personalization (β=0.39, p<0.001) and cultural relevance (β=0.31, p<0.001). The research proposes a design framework for AI-enabled science communication tools and highlights their role in bridging urban scientific literacy gaps.