The Shifting Dynamics of Enterprise AI: Balancing Urgency and Flexibility

The Evolution of AI in Business Strategy

Early in the enterprise AI race, urgency dominated the approach of many companies. Boards pressured leaders to keep pace, with markets expecting boosts in productivity from AI investments. The primary fear centered around being left behind. To meet these demands, leaders chose the most prominent AI providers, rapidly deploying solutions while sidelining deeper inquiries into their objectives. Although the pressure remains, its nature has evolved.

Changing Concerns Among CEOs

Recent research by AI platform Dataiku and The Harris Poll highlights a shift in the mindset of CEOs. About 65% now express concerns about over-investing in AI rather than under-investing. This marks a significant change from earlier stages. Revenue growth now slightly surpasses productivity as the main indicator of AI success, signaling boards’ expectations have shifted from mere experimentation.

The stakes for CEOs have intensified. A significant portion, 77%, predict that peers may be ousted due to failed AI strategies or crises driven by AI.

“CEO confidence in deploying AI fell even as the investment rose,” notes Florian Douetteau, Dataiku’s CEO and co-founder.

This sentiment arises from the realization that the more investment companies make, the more leaders recognize their lack of control over new systems.

Unanticipated Structural Risks

Many companies have entrenched themselves in vendor relationships that are challenging to exit. Pricing remains opaque, consumption patterns fluctuate, and technological capabilities evolve. Businesses that committed to one provider assumed stability in the relationship. Yet, unexpected changes like contract renewals, model deprecations, or superior competitor offerings can disrupt workflows. “It’s akin to pouring cement around the furniture,” Douetteau remarks, “only to discover the furniture will be moved multiple times before completing the house.”

Risks transcend commercial agreements. As AI infrastructure gains geopolitical sensitivity, regulations, export controls, or government actions can affect access, regardless of vendor contracts. While contracts govern pricing, service levels, and usage rights, they can’t shield companies from policy shifts impacting model accessibility and usage conditions.

A striking 76% of CEOs perceive significant operational or strategic risks from relying on too few AI vendors. Furthermore, 67% have questioned AI vendor decisions made by their teams within the past year. Meanwhile, 74% of IT decision-makers identify fragmented AI tools as substantial barriers to scaling efforts, according to a survey from Dataiku and Morning Consult.

As AI spreads throughout enterprises, decisions about its integration are scattered across various teams and vendors. Responsibility, however, largely resides with the CEO. Douetteau pinpoints a notable gap: while 70% of CEOs claim ownership of AI strategy, a mere 6% partake in daily decision-making. “This gap fosters dependency,” he states, “because the comprehensive view is never controlled by those observing its formation.”

Prioritizing Flexibility and Control

Rather than pursuing the ideal vendor, CEOs increasingly value adaptability over time as vendors and technologies evolve. Maintaining institutional knowledge and governance matters more than merely holding licenses.

A vendor relationship offers access as long as permitted, whereas an orchestration layer above any provider offers freedom to change models without rebuilding underlying work. This retains logic, governance, and knowledge under enterprise control.

Douetteau emphasizes, “We designed Dataiku to serve as this layer.” Dataiku provides a governed AI environment enabling teams to build, deploy, and adapt AI across existing vendors and models, preserving control and traceability.

Achieving this is crucial not only for operational efficiency. It impacts 81% of CEOs whose AI decisions presently shape their long-term legacies. Companies emerging strongest won’t be those moving swiftly but those establishing systems flexible enough to adapt to market shifts, ensuring continued understanding of operations.

“The question for any CEO is narrow,” Douetteau asserts. “Which elements of their company’s judgment are integrated into systems they control? Can they justify those systems’ actions to regulators or themselves?”

Successful enterprises won’t merely maintain flexibility; they will preserve control over judgment, governance, and workflows vital for AI operations.

For further insights into survey findings, review the Global AI Confessions Report: CEO Edition.

Research was conducted online by The Harris Poll on behalf of Dataiku (February–March 2026), surveying 900 CEOs at companies with annual revenue of $500M or more across the US, UK, France, Germany, UAE, Japan, South Korea, and Singapore.