Most companies still lag behind on AI targets—how can they catch Up? (www.techradar.com)

AI is widespread—78% of companies use it and the market tops $244B—but adoption isn’t translating into financial returns. CEOs report only 25% of AI initiatives met expected ROI and just ~18% cleared the cost of capital; IBM research finds one in three firms pause deployments after pilot. Common pitfalls are fragmented “hybrid-by-default” architectures, narrow data silos, aging compute/storage/network stacks (about 80% of orgs), and inadequate consideration of where models run—data centers, cloud, and edge—leading to limited scale, rising costs, and mounting security and compliance exposure. The stakes are high: the average U.S. data breach cost surpassed $10M in 2025 and 35% of breached firms faced regulatory fines over $250k. The path to closing the gap is technical and organizational: adopt a hybrid-by-design, platform-based AI stack that brings AI to the data and provides consistent capabilities across cloud, edge, and mainframe; modernize compute, storage, and networking for cost-performance balance; enforce end-to-end data governance, sovereignty, and privacy; and use common data science stacks to enable cross-team workflows (e.g., integrating chatbots with fraud, loans, and portfolio systems). Strategic partnerships and industry-wide collaboration further amplify scale and innovation. In short, intentional infrastructure, rigorous data practices, and coordinated ecosystems—not isolated pilots—turn AI projects into sustainable ROI.