Brief IA
›

AI Costs: Capgemini Advocates Measuring Created Value

Business & Startups·Tom Levy·

AI Costs: Capgemini Advocates Measuring Created Value

AI Costs: Capgemini Advocates Measuring Created Value
Key Takeaways
159% of executives surveyed by Capgemini report having exceeded their AI spending forecasts.
2For 20% of organizations, the budget gap exceeds 30%.
3Three out of five executives plan to adopt a multi-model strategy within the next 12 months.
Why it matters : The difficulty in linking AI usage to tangible business outcomes complicates the justification of expenses.

Capgemini observes that the majority of organizations struggle to link their AI usage to concrete business outcomes, complicating the justification for rising expenditures.

Budget Monitoring Expected to Drop to 44%

The Capgemini Research Institute forecasts a shift in AI management criteria: organizations are expected to place greater importance on business success indicators derived from AI. Currently, 60% of companies measure their total AI spending against the budget, but this proportion is expected to fall to 44% in the next 6 to 12 months.

Capgemini defines AI performance as an indicator that allows for the verification and recording of the value generated by AI. A more refined assessment of this performance could support requests for increased budgets dedicated to tokens. Volker Darius, CEO of Capgemini Invent, believes that the growing integration of AI into work processes necessitates managing it as a portfolio of economic outcomes, rather than merely as a technical or procurement issue.

59% of Executives Report Exceeding Their Forecasts

In September 2026, 500 CXO-level executives from large organizations were surveyed by Capgemini. Among them, 59% report that their AI spending has exceeded forecasts, and for 20% of organizations, the budget overrun exceeds 30%. These overruns are leading an increasing number of companies to involve CFOs more in AI-related spending decisions.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

More than 53% of executives express concern about AI costs, even as the price of tokens for the most popular models has decreased. This cost remains a source of worry, but training and governance also weigh heavily on the bill. Additionally, human oversight and repetitive task costs contribute to the uncertainty regarding the total expenditure.

Three in Five Executives Plan to Diversify Models

To address rising costs, Capgemini observes that companies are tightening their budgetary frameworks and diversifying the AI models used, an approach referred to as a multi-model strategy. Three in five executives surveyed plan to adopt this strategy in the next 12 months, selecting models based on their cost, capabilities, latency, and risk.

Volker Darius, CEO of Capgemini Invent, emphasizes that cost discipline alone is not sufficient to make AI profitable, and that creating value without spending control is not sustainable. He recommends measuring cost, quality, and business value together, choosing the most suitable model for each task, and considering the necessary human oversight to ensure safe outcomes. According to him, continuous monitoring of AI performance would better guide capital allocation decisions and develop AI with greater confidence.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.