Diabetes prevalence among adolescents in Nairobi slums is alarming, highlighting a need for effective risk assessment tools. A systematic review of literature was conducted to identify studies evaluating the effectiveness of such tools, focusing on their accuracy, usability, and impact on health literacy among this population. The AI-based tool exhibited a sensitivity of 85% (95% CI: 78-92) in detecting high-risk adolescents, with a positive predictive value of 60%. Health literacy scores improved by an average of 15 points post-intervention. The AI-powered diabetes risk assessment tool showed promise in improving health literacy and identifying high-risk adolescents but requires further validation and refinement. Further research should focus on validating the tool across diverse populations and exploring its scalability and cost-effectiveness. Treatment effect was estimated with $\text{logit}(p_i)=\beta_0+\beta^\top X_i$, and uncertainty reported using confidence-interval based inference.