2026年首席数据官 (CDO) 洞察报告：数据治理、数据信任悖论与AI素养成为核心焦点

CDO Insights 2026 Data governance and the trust paradox of data andAI literacy take center stage Contents Executive Summary3Key Findings4Making Room for AI5-Spotlight: Organizations Eye Customer Experience Enhancements with AI6Taking AI from Pilot to Production7-Spotlight: Improvements to Data and Resources Are Aiding GenAI Success8The Need to Temper Employee Trust9Governing AI for Confidence in Responsible Use10Data Management Investments12Vendor Support13Conclusion14About Wakefield Research15Methodology Notes15About Informatica16About Salesforce16About Deloitte16 Executive Summary The moment has come: AI has transitioned from experimental technology to a strategic, transformational imperative. Now comes the hard part. With organizations exploring use cases and steering initiatives toward production, many are discovering thatthe growing, widespread use of AI at their companies is exposing infrastructure and governance gaps as well asshortfalls in employees’ data literacy and AI literacy. The enthusiasm around AI is being tempered by the realizationthat there are high-stakes challenges ahead that need to be addressed if ongoing and future AI initiatives are tocontinue finding success. Many are embracing the mindset that going all-in on AI is the only way to make it work. AI adoption has reached a tipping point at companies, with 69% having incorporated generative AI (GenAI) into theirbusiness practices and another 25% expecting to do so in the next 12 months, according to a survey of 600 dataleaders (CDOs, CDAOs and CAOs) from companies with $500M+ in revenue across the U.S., UK / EU and APACconducted by Wakefield Research for Informatica. That’s a sharp rise from48% who had adopted GenAI last yearand45% who had adopted it the year prior.Agentic AI, seen by many as an important next step in AI use — moving from content creation to autonomousexecution and decision making — is following suit with nearly half (47%) already having adopted the technology. While there are numerous factors that can ultimately make or break the success of an AI initiative, not groundingthe initiative in high quality data and a strong underlying data management foundation can destroy a project beforeit takes off. The majority (61%) cite better data — including higher data quality and completeness — as a factorthat is making it easier to successfully transition more of their GenAI pilots into production. That said, the job is farfrom finished: among data leaders at companies that have adopted or plan to adopt agentic AI, half (50%) cite dataquality / retrieval concerns as a top challenge for moving AI agents into production. As AI initiatives spread, there’s another widespread concern on data leaders’ minds:75% believe their workforceneeds upskilling in data literacy and 74% in AI literacy to responsibly use AI or AI outputs in day-to-dayoperations. Within organizations, however, trust in the data being used to drive AI is high: 65% of data leadersbelieve most, all or almost all employees at their company trust the data they have and are using for AI.But thathigh trust is a cause for concern when there are underlying data reliability problems and data and AI literacygaps preventing workers from being able to recognize and address the potential shortcomings of the data beingused for their AI efforts. In addition,76% acknowledge their company’s visibility and governance have not completely kept up withemployees’ use of AI, which could increase exposure issues for companies already grappling with vulnerabilities. The great push to embrace the transformational possibilities of AI companywide increases the risks companiesare taking if they don’t have their data houses in order. As companies look to the future, 86% will increase datamanagement investments to help navigate the trust paradox and address issues ranging from data privacy andsecurity to AI visibility and governance as well as lagging employee data and AI literacy. The Current State of AI As adoption ramps up for agentic AI, there’s a splitamong companies who’ve adopted or plan to adoptagentic AI on how they’ll manage or deploy AI agents.For 54%, expectations are that it’ll be done with vendorsolutions while 44% expect to develop and manage AIagents internally. Making Room for AI Adoption is speeding up for GenAI, andagentic AI technologies are following suit GenAI adoption has jumped to 69%. In comparison,from two years ago to last year, those who saidtheir company had adopted GenAI into its businesspractices had only increased from 45% to 48%. Thatramp-up isn’t over. An additional 25% expect theircompany to adopt GenAI in the next 12 months, leavingjust 6% who don’t anticipate using GenAI by this timenext year. GenAI adoption is highest in the U.S. (75%,compared to 48% the year prior), followed by the UK /EU and APAC (both 66%, compared to 47% and 51% in2025, respectively). More than 2 in 3 havealready adopted GenAI And many are continuing this trend with a