Capgemini CEO Warns Against Rushing AI Adoption

Corporate leaders are increasingly plagued by a fear of missing out on applied artificial intelligence, driving them to make large capital expenditures despite unclear outcomes. Aiman Ezzat, the chief executive of Capgemini, warns that this anxiety is leading businesses to tread a dangerous line between over-investing and lagging behind. He suggests that the current rush is not a strategic necessity but a reaction to uncertainty, where companies are building capabilities that may not yet be wanted by the market.

Ezzat draws a parallel to the metaverse, a technology that once dominated boardroom conversations but has since faded into obscurity. He notes that Capgemini itself experimented with metaverse labs, only to recognize that the timing was off. The lesson, he argues, is that technology does not mature in a single dramatic moment. Instead, it evolves in increments, and companies that invest too heavily before the technology is ready risk wasting resources on solutions that no one is asking for.

Strategy requires measured investment

The recommended approach is agility rather than aggressive expansion. Ezzat emphasizes the need for small tests and pilots before scaling up, a strategy designed to keep companies aware of technological shifts without overcommitting. Capgemini maintains laboratories for emerging fields like quantum computing and 6G mobile technology, not to dominate these sectors immediately, but to be positioned to scale when adoption accelerates. This method allows firms to avoid the costly mistake of being too far ahead of the learning curve.

According to reporting by GN technics/ai (en-US), this cautious stance is a reaction to the volatility seen in the broader tech sector. Many companies are currently viewing AI primarily as a tool for internal efficiency, such as streamlining finance or human resources. However, Ezzat argues that this is a limited view. He insists that AI should be treated as a business transformation strategy rather than just a separate technology project. The goal is to connect disparate parts of the enterprise in innovative ways, rather than simply using AI to keep existing operations running.

Human trust remains the barrier

A significant challenge in this integration is the lack of trust between humans and AI agents. While the industry often uses the phrase "human in the loop," Ezzat suggests that the reality is more complex. He argues that the true issue is "AI-human-centricity," where the system must be designed to earn the trust of the human user. Currently, while an AI agent may be programmed to trust a human, the human does not reciprocate that trust, creating a friction point that hinders full adoption.

This perspective aligns with historical engineering principles from the ergonomics movement, which focused on designing tools for human comfort and capability rather than just industrial efficiency. Ezzat believes that modern AI integration must follow a similar path, ensuring that technology serves human needs and decision-making processes. Without this focus, companies risk creating systems that are technically advanced but practically unusable, leaving them stuck on the starting blocks while competitors find a more balanced approach.