Dataiku’s report highlights a significant disconnect in perceptions regarding AI’s reliability between CEOs and data leaders. The survey, conducted by Harris Poll across multiple countries, indicates that 95% of data leaders lack full visibility into the AI decision-making process. Although 86% of respondents recognized AI’s integration into daily operations, issues regarding governance gaps, lack of explainability, and excessive trust were prevalent. A mere 19% required AI agents to disclose their processes, while 52% experienced delays or halts in deployment due to explainability deficiencies.
The report shows increasing accountability for AI outcomes among CIOs and CDOs, with 46% stating they expect recognition for successful AI implementation, while more than half fear blame for failures. Additionally, 60% worry about job security if AI fails to deliver visible results in the next two years. Despite believing AI can outperform business analysis, 74% of respondents would revert to a human-centered approach if error rates exceed 6%. This reflection points to a broader concern about reliance on AI, with 89% of respondents citing at least one task they would never delegate to AI.
👉 Pročitaj original: CIO Magazine
