The priority ranking that should reorder your budget
Info-Tech Research Group's Future of IT survey, drawing on a base the firm says covers more than 30,000 IT, HR, and marketing professionals, found that 40.9% of leaders name improving data governance as a top data priority for 2026. The notable detail is that this ranks above AI-specific initiatives, in a year when AI dominates nearly every other enterprise technology conversation and budget line.
That ordering is a useful corrective for any organization that has been funding AI pilots and proof of concepts faster than it has funded the governance work underneath them. A survey base this large naming governance as the top concern, ahead of the technology everyone is excited about, suggests practitioners on the ground have already learned what a lot of executive roadmaps have not yet absorbed: the AI initiative is not the bottleneck, the data underneath it is.
Four priorities, and why the order matters
Info-Tech's report lays out four recommended data priorities for 2026. The first, enterprise-wide accountability, calls for a single governance framework covering both data and AI, replacing the common pattern where data governance and AI governance live in separate committees with separate owners and inconsistent rules. The second, customer-centricity for data, pushes organizations to treat data as a reusable, outcome-driven product rather than a byproduct of whatever system generated it.
The third and fourth priorities get less attention but matter just as much. Trusted AI-ready data supply means improving quality and transparency so analytics and AI can actually rely on what they are fed, rather than quietly absorbing bad inputs that surface later as bad outputs. Data champions, the fourth priority, is about building literacy and collaboration across the organization rather than concentrating data skill in a single specialist team that becomes a bottleneck for everyone else's AI ambitions.
The quote that captures the whole argument
Info-Tech senior research analyst Pooja Khandelwal frames the stakes in a line that deserves to be read twice: feed AI and automation noise, and they will scale confusion, feed them clarity, and they will scale intelligence. That is a precise description of why so many enterprise AI pilots produce underwhelming results despite using genuinely capable models. The model is not the limiting factor. The data feeding it is.
Khandelwal's second point reinforces the first: without these data foundations, even the most advanced AI and IT initiatives will stall or fail to deliver value. That is a direct challenge to any organization treating data governance as a compliance checkbox rather than a prerequisite for the AI returns the board is expecting this year.
Governance competing with budget pressure, not riding a tailwind
Khandelwal explicitly flags that turbulence in the economic and regulatory landscape will continue into 2026, which matters because it means governance investment has to win budget arguments against real cost pressure, not coast on enthusiasm the way AI line items currently do. A governance program is harder to fund than a flashy AI pilot precisely because its payoff is the absence of a future failure, not a demo that impresses a board member.
That dynamic puts CIOs and CDOs in an uncomfortable position: the investment their own survey data says practitioners want most is also the hardest one to defend in a budget review built around visible, demoable progress. Making the case requires tying governance spend directly to specific AI initiatives that are already stalled or underperforming because of data quality, rather than pitching governance as an abstract best practice.
What to actually do with this finding
If your organization's 2026 planning has AI initiatives funded ahead of, or separate from, data governance investment, this survey is a reason to revisit that sequencing before the budget locks in. The specific move Info-Tech recommends, a single governance framework spanning both data and AI rather than two parallel committees, is a concrete starting point that does not require a large new program, just consolidating ownership that is probably already fragmented across your organization.
The more useful internal exercise is auditing your current AI pilots against the four priorities directly: do they have a designated data product owner, a trust score or quality baseline, and someone accountable for governance who is not simply the data team hoping compliance notices later. A stalled AI pilot audited against this checklist usually reveals which of the four foundations was skipped, and that gap is almost always more fixable, and cheaper, than building a new model.
