AI, and generative AI in particular, offers organizations a way to get more from their computing infrastructure while simultaneously placing new demands on it. Among all applicants to the MINDS programme, “limitations of technical infrastructure” was cited among the top three challenges to achieving AI impact. Across the board, the focus is moving from simply expanding infrastructure to developing strategic engineering capabilities that create the right conditions for AI to deliver impact at scale. Unified AI platforms enable scalability and agility, unlike fragmented point solutionsMINDS organizations are rethinking the foundations of their technology to stay competitive and agile to evolving business needs. On the hardware side, they are expanding computing power, improving data storage and connecting edge devices that process information closer to where it is generated (such as in factories, hospitals or logistics centres). On the software side, they are building unified data environments that connect models, workflows and applications, supported by robust security and model-serving capabilities. Together, these integrated platforms become strategic enablers for AI, reaching far beyond what fragmented point solutions can offer (see Spotlight 7). By embedding AI into unified architectures, organizations can unlock interoperability and scalability, reduce redundancy and accelerate time-to-value in multiple applications. AI platforms also provide structural flexibility to adapt to evolving market conditions and regulatory landscapes. This positions organizations not only to optimize current processes but also to lead in shaping future industry standards and impacting sustainable growth in an AI-driven economy.2.4 Insight 4: Modernizing the technology stack for advanced AI capabilities SPOTLIGHT 7 State Grid Corporation of China’s platform for city-scale intelligent energy management in Shanghai Shanghai’s rapid growth has created immense pressure on its power grid, which serves more than 12 million residential and corporate users with increasingly fragmented and renewable-heavy energy profiles. Traditional forecasting and trading systems cannot keep pace with this complexity. To solve this, State Grid Corporation of China launched the Intelligent Energy Management Master, a city-scale AI platform that uses orchestration rather than isolated tools. The platform integrates four intelligent agents for forecasting, trading, regulation and settlement within a unified system used by grid operators and energy managers. Generative AI optimizes decisions under tight constraints, while human- in-the-loop oversight ensures trust and scalability through intuitive dashboards that visualize real-time data and resource flows. Impact: As a result of its platform-centric AI strategy, forecasting accuracy improved by 12.5%, grid reliability reached 99.9983%, and the system saved over $1.12 billion by avoiding costly new generation and transmission investments. Unified platforms for building efficiency and quality controls BOX 14 Siemens applies a platform-based AI approach across multiple domains, including building efficiency and comfort (where AI-based closed- loop HVAC optimization improves comfort compliance by 28% while reducing monthly energy consumption) and quality control (where Siemens standardized their inspections with EthonAI’s industrial AI platform). The latter, integrated with Siemens Industrial Edge, enables factories across Europe, North America and Asia to automate millions of visual inspections and uncover inefficiencies through causal AI. Standardizing visual inspection via a unified platform – rather than isolated point solutions – enables savings between $34,000–115,000 per inspection station deployed. MINDS organizations are rethinking the foundations of their technology to stay competitive and agile to evolving business needs. Proof over Promise: Insights on Real-World AI Adoption from 2025 MINDS Organizations 20

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