The Chief Data Officer (CDO) has evolved from a typical compliance role to one of the most critical positions for AI deployment. These executives now sit at the intersection of data governance, AI strategy, and workforce readiness. Their decisions determine whether enterprises move from AI pilot to production scale or remain stuck in experiment mode.
That’s why Informatica’s third annual survey – the largest ever survey of CDOs on AI readiness, involving 600 executives globally – holds special significance. The findings highlight a dangerous disconnect that explains why so many organizations struggle to scale AI beyond pilots: While 69% of enterprises have deployed generative AI and 47% are running agentic AI systems, 76% admit that their governance frameworks may not keep pace with how employees actually use these technologies.
Survey shows what Informatica says "belief paradox" – and explains why data leaders are dangerously overconfident about AI readiness. Organizations deployed generic AI systems faster than they could build the governance and training infrastructure to support them. The result: Employees generally trust data powering AI systems, but organizations acknowledge that their workforce lacks the literacy to question that data or use AI responsibly. Seventy-five percent of data leaders say employees need to increase skills in data literacy. Seventy-four percent require AI literacy training for day-to-day tasks.
"The only difference now is can you trust the data to upset an agent?" Informatica CIO Graeme Thompson told VentureBeat. "Agents do what they’re supposed to do if you give them the right information. There is such a lack of confidence in the data that I think that’s the difference."
Why isn’t infrastructure a barrier to data and AI?
GenAI adoption has increased from 48% a year ago to 69% today. Nearly half of organizations (47%) now run agentic AI – systems that take actions autonomously rather than simply generating content. This rapid expansion has created a race to acquire vector databases, upgrade data pipelines, and expand compute infrastructure.
But Thompson dismisses the lack of infrastructure as the primary problem. The technology exists and works. The limitation is organizational, not technical.
"The technology, the infrastructure that we have available at the moment is much more than that – that’s not the problem right now," Thompson said. He compared the situation to amateur athletes blaming their equipment. "There’s still a long way to go before the equipment in the room becomes a problem. People chase equipment like golf players. Those golfers are a sucker for a new driver, a new putter that is going to fix their physical inability to hit the golf ball straight."
Survey data supports this. When asked about 2026 investment priorities, the top three are all people and process issues: data privacy and security (43%), AI governance (41%), and workforce upskilling (39%).
Five tough lessons for enterprise CDOs
The survey data combined with Thompson’s implementation experience shows specific lessons for data leaders trying to move from pilot to production.
Stop chasing infrastructure, solve people’s problems
The trust paradox exists because organizations can deploy AI technology faster than they can train people to use it responsibly. Seventy-five percent need data literacy upskilling. Seventy-four percent need AI literacy training. The difference of technology is the difference of people.
"Bringing in your people who know your company, know your data and know your processes to learn AI is much easier than bringing in an AI person and teaching them about your company who doesn’t know anything about those things." Thompson said. "And AI people are also very expensive, just like data scientists are very expensive."
Make CDOs an execution task, not an ivory tower
Thompson structures Informatica so that the CDO reports directly to him as the CIO. This makes data governance an execution function rather than a separate strategic layer.
"This is a deliberate decision based on a function that gets the job done rather than an ivory tower function," Thompson said. The structure ensures that data teams and application owners share common priorities through a common boss. "If they have a common boss, their priorities should align. And if not, it’s because the boss isn’t doing his or her job, not because both jobs aren’t working according to the same priority list."
If 76% of organizations cannot effectively control their use of AI, reporting structures may be part of the problem. Siled data and IT functions create conditions for pilots that never scale.
Build literacy outside of IT teams
The crucial insight is that AI literacy programs must extend beyond technology teams to business functions. At Informatica, the chief marketing officer is one of Thompson’s strongest AI partners.
"You need literacy in your business teams as well as your technology teams," Thompson said.
He said the marketing operations team understands technology and data. knows that the answer "How can I get more value from my limited marketing program dollars each year?" That happens by automating that work and adding AI, not by adding people and more Google ad dollars.
Business-side literacy creates a pull rather than a push for AI adoption. Marketing, sales, and operations teams begin to demand AI capabilities as they see strategic value, not just efficiency gains.
Treat AI as a strategic expansion, not a cost reduction
Data leaders have spent decades fighting the perception that IT is just a cost center. AI offers an opportunity to change that narrative, but only if CDOs redefine the value proposition away from productivity savings.
"I’m very frustrated that, given this new technology capability, as IT people and as data people, we immediately turn around and talk about productivity savings," Thompson said. "What a waste of opportunity."
Strategic shift: Pitch AI’s ability to completely remove headcount constraints rather than simply reduce existing headcount. This transforms AI from operational efficiency to strategic capability. Organizations can expand market reach, enter new geographic areas and test initiatives that were previously cost-prohibitive.
"It’s not about saving money," Thompson said. "And if that’s primarily your approach, your company is not going to win."
Go vertical first, scale the pattern
Don’t wait for the right horizontal data governance layers to be in place before providing production value. Choose a high-value use case. Build the entire governance, data quality and literacy stack for that specific workflow. Validate the results. Then repeat the pattern for adjacent use cases.
It provides output value by gradually building organizational capacity.
“I think this space is moving so fast that if you try to solve your governance problem 100% before you get to your semantic layer problem, before you get to your vocabulary problem, you’ll never get any results and people will lose patience," Thompson said.
<a href