This study reports an educational innovation in MA1053: Strategic Mathematics for a Sustainable Future (Digital Education, Tecnológico de Monterrey; Feb-July and Summer 2025). The course integrates algebra, calculus, and linear regression with AI, Python, and Excel to analyze PM2.5, aligned with the SDGs. Using a single-group pre-post design (n= 60), we assessed domain-specific cognitive performance (0 100) and procedural/ethical transfer via rubric (0-4 rightarrow 0-100). Significant gains were observed (¿=.05, Holm): Algebra ¿=8.54 [7.72, 9.36], d_z=2.69; Calculus ¿=12.04 [11.02, 13.05], d_z=3.06; Regression ¿=9.75[8.99,10.51], d z=3.31; overall ¿=10.11[9.56, 10.65], d_z=4.79. Project-rubric means increased from 50.26 to 68.02(¿=17.76; one-tailed Wilcoxon post > pre, p<.001), with the largest improvements in Modeling and Visualization/Regression. A 3 × 3 heatmap of correlations among ¿ indicated modest crossdomain coherence, consistent with area-specific skill growth. We conclude that integrating AI and emerging technologies can enhance mathematical learning and sustainability-oriented formation. Strengthening cross-domain 'bridging' tasks may further generalize gains. The main limitation is the single-cohort design without a control group; future work should extend to multiple sites/cohorts and track vertical transfer.
