95% of Data Leaders Admit They Can’t Fully Trace AI Decisions, Reveals New Dataiku “Global AI Confessions Report”

Findings from the Dataiku/Harris Poll survey of 800+ data leaders reveal 75% worry about trust in AI agent deployments

NEW YORK & SAN FRANCISCO, Oct. 07, 2025 (GLOBE NEWSWIRE) --

Survey Highlights

80% say an accurate but unexplainable AI decision is riskier than a wrong but explainable one.
69% report that AI business suggestions are taken more seriously than human ones.
Only 19% of data leaders always require AI agents to “show their work” before approval.
52% have delayed or completely blocked agent deployments due to explainability concerns.
73% say the C-suite underestimates the difficulty of achieving AI reliability prior to production.
58% of data leaders worry AI-generated code vulnerabilities are a “disaster waiting to happen.”

Despite rapid adoption, most data leaders say AI trust and explainability are major risks—exposing a dangerous gap between CEO ambition and enterprise reality.

Nearly all data leaders (95%) admit they lack full visibility into AI decision-making, according to the “Global AI Confessions Report: Data Leaders Edition,” from Dataiku, The Universal AI Platform™. The study was conducted by The Harris Poll for Dataiku, surveying 800+ senior data executives across the U.S., U.K., France, Germany, UAE, Japan, Singapore, and South Korea. The report reveals that, while 86% of respondents estimate AI has become embedded into their daily operations, concerns about governance gaps, lack of explainability, and misplaced confidence remain. Surprisingly, only 19% of data leaders always require AI agents to “show their work” before approval, and 52% have delayed or completely blocked agent deployments due to explainability concerns, specifically.

The Pressure on Data Leaders

CIOs and CDOs carry the weight of responsibility: 46% are most likely to be credited for AI gains, but over half, 56%, are most likely to be blamed for business losses due to failed AI. With 60% fearing their own jobs are at risk if AI doesn’t deliver measurable results within two years, the stakes for data leaders have never been higher.

Fragile Trust in AI Outcomes

Despite optimism about potential, AI is already causing headaches for global enterprises:

59% report that AI hallucinations or inaccuracies have already caused business issues in the past year.
82% believe AI can outperform their boss in business analysis, but 74% would revert to human-managed processes if AI error exceeds just 6%.
89% say there is at least one business function they would never delegate to AI.

The CEO Disconnect: Why AI Is Stuck in Proof-of-Concept (POC)

Earlier this year, Dataiku's "Global AI Confessions Report: CEO Edition" showed how bullish CEOs are on AI. Data leaders, however, are more skeptical. Only 39% say their C-suite truly understands AI, 68% believe executives overestimate its accuracy, and 73% say they underestimate the difficulty of achieving AI reliability prior to production.

The stakes are high, with 56% of data leaders expecting a CEO will be ousted by 2026 due to a failed AI strategy. This disconnect, CEO optimism versus data leader caution in pushing unfit AI into production, could explain why so many projects remain stuck in POC.

Bridging the Gap

“An alarming revelation of the report is that enterprises worldwide are betting on AI they don’t fully trust. The good news is that most failed AI initiatives suffer from common blockers that can be overcome with more explainability, traceability, and governance. That’s how AI moves from hype to real business impact,” explained Florian Douetteau, Co-founder and CEO of Dataiku.

The full report, “Global AI Confessions Report: Data Leaders Edition”, is available here: [Download Report]

Harris Poll Survey Methodology

The research was conducted online by The Harris Poll on behalf of Dataiku from August 20 –29, 2025. The survey was conducted among professionals considered “Data Leaders” in the United States, the United Kingdom, France, Germany, the UAE, Japan, Singapore, and South Korea. A total of 812 interviews were conducted (USA = 203, UK = 102, France = 101, Germany = 103, UAE = 100, Japan =103, South Korea = 50, Singapore = 50). Data Leaders in the study work for large companies with an annual revenue of or more than $1 Billion or regional equivalents, and are defined by professional titles ranging from vice-president, director, managing director, or C-suite level.

About Dataiku

Dataiku is The Universal AI Platform™, accelerating the delivery of enterprise AI solutions directly tied to business efficiency and revenue growth. Running on top of any data platform and any LLM, it unites all teams, from business users to data scientists, and all AI techniques, from predictive analytics to agentic, in one secure, governed environment.

Dataiku is trusted by 1 in 4 of the world’s top companies, and is backed by investors, including Wellington Management, Battery, CapitalG, ICONIQ, and FirstMark. For more, visit the Dataiku blog, LinkedIn, X, and YouTube.

Kevin McLaughlin
Dataiku
press@dataiku.com

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.