The chief information officer (CDO) has developed from a distinct segment compliance function into one in every of the most important positions for AI deployment. These executives now sit at the intersection of knowledge governance, AI technique, and workforce readiness. Their choices decide whether or not enterprises transfer from AI pilots to manufacturing scale or stay caught in experimentation mode.
That is why Informatica’s third annual survey — the largest survey yet of CDOs specifically on AI readiness, spanning 600 executives globally — carries specific weight. The findings expose a harmful disconnect that explains why so many organizations wrestle to scale AI past pilots: Whereas 69% of enterprises have deployed generative AI and 47% are operating agentic AI programs, 76% admit their governance frameworks cannot maintain tempo with how staff really use these applied sciences.
The survey reveals what Informatica calls a “belief paradox” — and explains why information leaders are dangerously overconfident about AI readiness. Organizations deployed generative AI programs sooner than they constructed the governance and coaching infrastructure to help them. The outcome: Staff typically belief the information powering AI programs, however organizations acknowledge their workforces lack the literacy to query that information or use AI responsibly. Seventy-five % of knowledge leaders say staff want upskilling in information literacy. Seventy-four % require AI literacy coaching for day-to-day operations.
“The hole now is simply, are you able to belief the information to set an agent free on it?” Graeme Thompson, CIO at Informatica, informed VentureBeat. “The brokers do what they’re supposed to do should you give them the proper information. There’s simply such an absence of belief in the information that I feel that is the hole.”
Why infrastructure is not the bottleneck for information and AI
GenAI adoption jumped from 48% a yr in the past to 69% at this time. Practically half of organizations (47%) now run agentic AI — programs that autonomously take actions relatively than simply generate content material. This fast enlargement has created a race to purchase vector databases, improve information pipelines, and increase compute infrastructure.
However Thompson dismisses infrastructure gaps as the major drawback. The expertise exists and works. The limitation is organizational, not technical.
“The expertise that we now have accessible at the second, the infrastructure, is greater than — it is not the drawback but,” Thompson mentioned. He in contrast the state of affairs to newbie athletes blaming their tools. “There is a great distance to go before the tools is the drawback in the room. Folks chase tools like golfers. These golfers are a sucker for a brand new driver, a brand new putter that is going to treatment their bodily lack of ability to hit a golf ball straight.”
The survey information helps this. When requested about 2026 funding priorities, the prime three are all folks and course of points: information privateness and safety (43%), AI governance (41%), and workforce upskilling (39%).
5 laborious classes for enterprise CDOs
The survey information mixed with Thompson’s implementation expertise reveals particular classes for information leaders attempting to transfer from pilots to manufacturing.
Cease chasing infrastructure, repair the folks drawback
The belief paradox exists as a result of organizations can deploy AI expertise sooner than they’ll prepare folks to use it responsibly. Seventy-five % want information literacy upskilling. Seventy-four % want AI literacy coaching. The expertise hole is a folks hole.
“It is a lot simpler to get your people who know your organization and know your information and know your processes to study AI than it is to convey an AI particular person in that does not know something about these issues and educate them about your organization,” Thompson mentioned. “And in addition the AI folks are tremendous costly, similar to information scientists are tremendous costly.”
Make the CDO an execution operate, not an ivory tower
Thompson constructions Informatica so the CDO experiences instantly to him as CIO. This makes information governance an execution operate relatively than a separate strategic layer.
“That is a deliberate determination based mostly on that operate being a get issues completed operate as a substitute of an ivory tower operate,” Thompson mentioned. The construction ensures information groups and utility homeowners share frequent priorities by way of a standard boss. “If they’ve a standard boss, their priorities ought to be aligned. And if not, it is as a result of the boss is not doing his job, not as a result of the two features aren’t working off the identical precedence listing.”
If 76% of organizations cannot govern AI utilization successfully, reporting construction could also be a part of the drawback. Siloed information and IT features create the situations for pilots that by no means scale.
Construct literacy outdoors IT groups
The breakthrough perception is that AI literacy packages should prolong past expertise groups into enterprise features. At Informatica, the chief advertising officer is one in every of Thompson’s strongest AI companions.
“You want that literacy throughout your enterprise groups in addition to in your expertise groups,” Thompson mentioned.
He famous that the advertising operations group understands the expertise and information. It is aware of that the reply to the “How do I get extra worth out of my restricted advertising program {dollars} annually?” is by automating and including AI to how that job is completed, not including folks and extra Google advert {dollars}.
Enterprise-side literacy creates pull relatively than push for AI adoption. Advertising and marketing, gross sales and operations groups begin demanding AI capabilities as a result of they see strategic worth, not simply effectivity positive aspects.
Pitch AI as strategic enlargement, not value discount
Information leaders have spent many years preventing perceptions that IT is only a value middle. AI presents the alternative to change that narrative, however provided that CDOs reframe the worth proposition away from productiveness financial savings.
“I’m very upset that, given this new expertise functionality on a plate, as IT folks and as information folks, we instantly flip round and speak about productiveness financial savings,” Thompson mentioned. “What a waste of a chance.”
The tactical shift: Pitch AI’s capacity to take away headcount constraints totally relatively than cut back present headcount. This reframes AI from operational effectivity to strategic functionality. Organizations can increase market attain, enter new geographies and take a look at initiatives that have been beforehand cost-prohibitive.
“It is not about saving cash,” Thompson mentioned. “And if that is primarily the strategy that you’ve, then your organization’s not going to win.”
Go vertical first, scale the sample
Do not await excellent horizontal information governance layers before delivering manufacturing worth. Choose one high-value use case. Construct the full governance, information high quality and literacy stack for that particular workflow. Validate outcomes. Then replicate the sample to adjoining use instances.
This delivers manufacturing worth whereas constructing organizational functionality incrementally.
“I feel this house is transferring so shortly that should you attempt to clear up 100% your governance drawback before you get to your semantic layer drawback, before you get to your glossary of phrases drawback, then you definately’re by no means going to generate any final result and folks are going to lose persistence,” Thompson mentioned.
Disclaimer: This article is sourced from external platforms. OverBeta has not independently verified the information. Readers are advised to verify details before relying on them.
