Capgemini bets on 'modernization supercycle' as AI forces IT overhaul

Capgemini raised its 2026 revenue growth target after stronger bookings, with CEO Aiman Ezzat telling analysts that companies chasing AI at scale will first spend years overhauling legacy technology. The obstacle to enterprise AI is not model access, he said. It is fragmented data, decades of technical debt, and infrastructure never built for automation.

Speaking after the earnings release on July 30, Ezzat framed the situation bluntly: "Every organization today wants to become agentic. But before they can become agentic, they must become AI-ready, and most are not." The firm sees a "multi-year modernization supercycle" as enterprises upgrade data platforms, applications, and core infrastructure before deploying AI across operations.

Why legacy systems block AI, not model access

Ezzat's argument inverts the usual AI narrative. Executives and vendors talk about models, benchmarks, and inference costs. Capgemini's view is that none of that matters if an organization's data sits scattered across incompatible systems and its workflows run on brittle, decades-old code.

Generative AI can produce answers. But performing business processes consistently, executing multi-step tasks, acting as an agent across departments, requires something the demos skip: a connected technology estate. When systems remain disconnected, AI tools cannot access reliable information or execute tasks reliably.

This is a self-serving argument for a consultancy that sells exactly those modernization projects. It is also, by most enterprise IT accounts, accurate. Technical debt accumulates silently until something exposes it. AI is exposing it at scale.

Spending shifts from pilots to transformation

Ezzat noted that enterprise AI spending is becoming "more targeted." Clients are moving away from standalone experiments and pilot projects toward large-scale transformation programs. The implication: the proof-of-concept phase is ending. Companies now want production systems, and production systems require the foundation work first.

“AI is not only creating demand for new business capability; it's also accelerating the modernization of the technology foundation on which those capabilities depend.”

— Aiman Ezzat, CEO, Capgemini

That acceleration benefits Capgemini directly. The company generates revenue from both sides: the AI workloads clients want and the infrastructure upgrades they need before they can run them. The raised guidance suggests bookings for both are coming in.

What this means for IT buyers

For CTOs and enterprise architects, Capgemini's framing validates a familiar headache. AI vendor pitches assume clean APIs, unified data, and flexible compute. The reality inside most large organizations is none of those things. Budgets for AI often compete with budgets for the plumbing AI needs.

Ezzat's comments suggest that competition is resolving in favor of the plumbing, at least for now. Enterprises are funding foundational work they deferred for years, reframed as AI enablement. Whether that reframing survives the next budget cycle depends on whether the AI capabilities actually arrive.

The open question

Capgemini did not disclose specific booking figures or break out AI-related revenue. The raised guidance signals confidence, but the supercycle thesis will be tested over multiple quarters. If enterprises complete their modernization projects and AI delivery stalls anyway, the narrative shifts. For now, the firm is betting that the foundation work will take years, and that it will be paid to do it.

Source: Tech-Economic Times / ET