The AI argument has shifted. Most organisations have already said “yes” to AI — the real question now is why the results feel so patchy. New tools are deployed, pilots are running and budgets keep rising, yet meaningful returns remain frustratingly inconsistent.
According to Cloudflare’s 2026 App Innovation Report, the problem isn’t the models. It’s the software underneath them.
Based on a survey of 2,300+ senior leaders across APAC, EMEA and the Americas, the report finds that application modernisation is the strongest predictor of AI success. Organisations ahead of schedule on modernisation are nearly three times more likely to see clear value from AI. In APAC, 92% of leaders say updating applications is the single biggest factor behind better AI performance.
The takeaway is blunt: AI is a foundation’s problem. Models need fast data access, flexible architectures and clean integrations. Legacy apps, brittle workflows and fragmented infrastructure choke progress, keeping AI stuck in demos and one-off use cases. Modernised systems, by contrast, make it easier to scale, iterate and actually ship.
Leading organisations are already moving beyond experimentation. In APAC, 90% have integrated AI directly into existing applications, with around 80% planning to go further over the next year. AI is becoming part of day-to-day systems — improving internal workflows, powering content-heavy apps and supporting revenue — rather than living in isolated pilots.
Lagging organisations pay the price in security and confidence. They tend to modernise reactively, often after incidents, and report lower trust in both their infrastructure and their teams. Poor alignment between security and application teams slows everything down, while technical debt and tool sprawl drain developer time.
The best performers are doing the opposite: cutting redundant tools, baking security into application design, and giving developers stable platforms to build on. The result is faster AI integration and fewer production headaches.
Cloudflare’s conclusion is clear. Winning with AI isn’t about chasing the latest model — it’s about having applications ready to use it. Without modern foundations, AI investment delivers shallow results. With them, AI finally starts to pay off.
Author: Mohammed Najem