Proof over Promise Insights on Real World AI Adoption from 2025 MINDS Organizations 2026

Among the MINDS organizations, the strongest results appeared when organizations’ efforts aligned with more than one of the five key insights shared above. These organizations demonstrated that the insights are not merely additive, but work cumulatively to create a multiplying effect, amplifying impact across business, society and the environment (see Spotlight 10).2.6 Amplifying impact through the interplay of key insights SPOTLIGHT 10 Foxconn’s transformation for self-learning factories Foxconn’s Project Genesis, developed in partnership with Boston Consulting Group (BCG), illustrates the combinatorial effect at work: the interplay between robust AI infrastructure, codified human expertise and a long-term vision guided by AI agents. Its goal is to redefine electronics manufacturing with AI infrastructure designed for autonomous, self-evolving factories. Overall, six breakthrough AI applications provide the foundation for the company’s next wave of digital transformation. Multiple agents enable shopfloor managers to monitor capacity, maintenance and layout optimization with precision and speed, moving operations from intelligence to autonomy and from experience-driven workflows to continuously learning systems.Crucially, Foxconn has digitalized decades of manufacturing know-how from experienced masters and paired technology with a mindset shift across global teams so that GenAI complements rather than replaces expert work. As Foxconn notes, “AI can power production, people power transformation”. Impact: 50% reduction in workload related to changeover; 30% decrease in problem resolution time; approximate 10% decrease in cycle time A unified approach to retail operations BOX 18 Sovereign LLM for cultural integration BOX 19Wumart and Dmall’s efforts to embed AI as a core capability within the company transformed retail operations by pairing optimal AI infrastructure with real-time data and scenario libraries for continuous improvement. Their approach moved the organization from digitally enabled stores to AI- orchestrated operations, addressing fragmented tools, manual routines and inconsistent execution. This resulted in significant daily profit increases, loss prevention and energy savings of 26%, demonstrating that infrastructure modernization, data strategy and workforce engagement are most powerful when pursued in tandem. Tech Mahindra’s approach to scaling language- inclusive AI models for public services is another powerful illustration of the combinatorial effect. By designing a multilingual, multimodal large language model tailored for low-resource languages and dialects, Tech Mahindra simultaneously tackled challenges in data diversity, technology stack modernization and responsible AI. Their system was trained using reinforcement learning with native speakers, ensuring contextual relevance and cultural alignment, while its modular architecture supported extensibility and data sovereignty. This strategy not only advanced inclusion and accessibility in citizen services, banking and healthcare, but also demonstrated how combining data strategy, infrastructure and responsible innovation can drive equitable and scalable impact. Proof over Promise: Insights on Real-World AI Adoption from 2025 MINDS Organizations 24