CEOs Claim AI Strategy Ownership, But CIOs Bear the Burden: Study Reveals Accountability Gap

Top Executives Say They’re in Charge—But Operational Reality Tells a Different Story

March 1, 2025 — A new global survey reveals a widening disconnect between who owns AI strategy and who executes it, raising concerns about corporate accountability and risk management.

According to Dataiku’s “Global AI Confessions Report; CEO Edition 2026,” commissioned via a Harris Poll of 900 enterprise CEOs worldwide, an overwhelming majority of chief executives assert that they personally direct their company’s AI strategy. However, the same executives acknowledge that day-to-day decision-making authority—and the associated liability—falls squarely on their chief information officers (CIOs).

“CEOs are under immense pressure to deliver measurable AI outcomes,” said Dr. Elena Marchetti, Dataiku’s chief AI ethics officer, in a statement accompanying the report. “Our research shows that while nearly 80% of CEOs claim full ownership of AI direction, only about 30% say they are directly involved in specific AI deployment decisions. That gap is where risk builds up.”

Boards demand progress; investors seek proof; markets require results. The survey indicates that many top leaders respond by publicly claiming clear ownership of AI strategy, yet the heavy lifting—and the potential for costly mistakes—remains with technology chiefs.

Background: The Pressure Cooker of AI Leadership

The study surveyed CEOs across industries in North America, Europe, and Asia-Pacific throughout late 2024. It found that 86% of respondents listed “AI strategy” as a top-three personal priority. Yet 62% admitted they do not feel fully equipped to evaluate AI project outcomes.

“This is a classic accountability gap,” said Raj Patel, a fellow at the Center for Digital Ethics and a former CIO. “CEOs own the vision; CIOs own the consequences. When an AI initiative fails—whether due to bias, regulatory violation, or cost overruns—the public often blames the technology leader, not the strategist.”

The report further highlights a mismatch in resource allocation: while 94% of CEOs stated they have a dedicated AI budget, only 44% confirmed that their organization employs a formal AI governance framework.

What This Means: A Risk Cluster Waiting to Ignite

Without a clear alignment between stated ownership and actual execution, companies face heightened exposure to compliance failures, ethical missteps, and financial losses. Regulators worldwide, including the European Union’s AI Office and the U.S. Federal Trade Commission, are increasingly targeting executive accountability for algorithmic decisions.

“If the CEO says they own AI strategy but the CIO makes every call on model deployment and monitoring, then when a regulatory inquiry happens, you have a finger-pointing exercise,” warned Patel. “The board won’t accept ‘the CIO did it’ as an answer.”

The report recommends that organizations establish explicit, written AI governance roles that bridge strategy and execution. It also suggests quarterly cross-functional audits to ensure that the CEO’s strategic vision translates into safe, accountable operational decisions.

As enterprises race to deploy generative AI and automated decision systems, the Dataiku study serves as a warning: declaring ownership is not the same as taking responsibility. The gap must be closed before the next headline-grabbing AI failure forces a reckoning.

↩ Read background | ↩ What this means