Dubai, UAE — A growing number of chief executives in the United Arab Emirates believe their tenure and long-term leadership prospects now depend on their ability to deliver measurable business outcomes from artificial intelligence initiatives, according to a study by Dataiku.
The findings, released in the 2026 edition of Dataiku’s CEO Confessions Study conducted by The Harris Poll, showed that 79% of UAE CEOs believe their role is at risk if their organisations fail to generate tangible business gains from AI by the end of 2026.
The study also found that 53% of UAE CEOs expect successful AI leadership experience to become the primary criterion boards use when appointing future chief executives within the next two years.
Dataiku said the UAE ranked highest globally for the proportion of CEOs who believe current AI decisions could damage their long-term legacy, with 23% expressing concern, more than double the global average.
The research suggests AI is increasingly being viewed not as a separate technology initiative but as a direct leadership responsibility within UAE boardrooms.
Executives take greater ownership of AI strategies
The study found that UAE chief executives are becoming more directly involved in AI decision-making as expectations from boards and investors intensify.
According to the report, 75% of UAE CEOs said their involvement in AI-related decisions had increased during the past year, while 55% identified themselves as the most influential stakeholder shaping organisational AI direction, ahead of information technology, data and business leaders.
Nearly six in ten CEOs in the UAE said they face pressure from boards to deliver measurable AI outcomes, although 90% described those expectations as realistic.
The report also found that 76% of respondents believe AI strategy and execution are important to investors, while 75% said CEOs could lose their jobs in 2026 because of a failed AI strategy or a high-profile AI-related crisis.
Despite mounting pressure to accelerate adoption, many companies remain cautious. Dataiku said 44% of UAE CEOs reported delaying or cancelling AI initiatives because of concerns over possible failure.
Florian Douetteau, chief executive and co-founder of Dataiku, said companies increasingly face pressure to translate AI investments into dependable business decisions.
“Every enterprise now has access to powerful AI. The differentiator is whether they can turn that power into reliable business decisions,” Douetteau said in a statement.
Governance concerns expose confidence gaps in organizations
The research also highlighted tensions between executive confidence in AI governance and concerns over accountability, transparency and regulatory scrutiny.
While 73% of UAE CEOs said they trusted their governance frameworks even if their jobs depended on them, the UAE ranked lowest globally in confidence in explaining AI-driven decisions to regulators or courts, according to the study.
The report further showed that 41% of CEOs had not challenged decisions made by AI vendors or platform providers within their organisations over the past year.
Dataiku said these findings reflected growing concerns over whether organisations possess sufficient oversight and flexibility to manage increasingly complex AI systems.
The study also found that 43% of UAE CEOs believe their organisations would face significant risk if the “AI bubble” were to burst, underlining concerns about long-term resilience alongside performance pressures.
Sid Bhatia, area vice president and general manager for the Middle East, Turkey and Africa at Dataiku, said governance was becoming central to sustaining trust and adaptability in enterprise AI systems.
“The CEOs who succeed are those who treat governance as an accelerator, not a constraint,” Bhatia said in a statement.
The online survey was conducted by The Harris Poll between Feb. 2 and March 2, 2026, among 900 CEOs from the United States, Britain, France, Germany, the UAE, Japan, South Korea and Singapore. Respondents led companies with annual revenues exceeding $500 million or regional equivalents.
