A Conference That Stopped Selling Optimism
From June 9 to 11, 2026, Info-Tech Research Group gathered thousands of CIOs and senior IT leaders at the Bellagio in Las Vegas for Info-Tech LIVE 2026, and the framing was notably less celebratory than the genre usually allows. CEO Tom Zehren challenged attendees to move beyond AI hype and focus on the operating discipline required to convert AI demand into measurable enterprise value. That is a pointed thing to say to a room that has spent two years being told AI changes everything. The pivot from possibility to discipline is the story of enterprise AI in 2026, and Info-Tech put it at the center of the stage.
The sentiment data explains why the tone shifted. Ninety-one percent of IT executives are bullish on AI and 96 percent expect their AI budgets to increase over the next twelve months, which on its own would read as unbroken momentum. The problem surfaces in the execution numbers. Only 42 percent of organizations report cross-departmental AI adoption with measurable impact, and only 50 percent have a board-approved dedicated AI strategy. Conviction and funding are nearly universal. Governance and proof are running at roughly half that rate, and the gap between the two is where value leaks out.
The Half That Has No Approved Strategy
We keep returning to the figure that half of organizations lack a board-approved AI strategy, because it sits uneasily next to nearly universal budget increases. It means a large share of enterprises are growing AI spend without a governing mandate from the body that approves the budget, which is governance by momentum rather than by design. In any other category of capital allocation that would draw scrutiny. In AI it has somehow become normal, and Info-Tech's data suggests the normalization is exactly the problem leaders should be unwinding rather than tolerating as a phase.
Chief Research Officer Gord Harrison tied the diagnosis to a prescription: "Agentic IT requires a different operating discipline that connects value creation and control through stronger strategy, governance, data readiness, security, and measurement." The phrase that earns its place is connects value creation and control. Most enterprises have treated those as opposing forces, where every control is a drag on value. Harrison's argument, echoed across the analyst community this quarter, is that in an agentic environment control is what makes value durable, because ungoverned agents produce gains that evaporate the moment one of them touches the wrong data.
Data Governance Is the Bottleneck
The most concrete finding from Info-Tech's diagnostics is that data governance is the single largest capability gap across its IT Management and Governance assessment, showing a 2.8-point spread between how important leaders rate it and how effective they are at it. That is the widest gap in the framework, and it is not an accident that it sits underneath the agentic ambitions everyone is funding. Agents are only as trustworthy as the data they act on, and a 2.8-point shortfall in the foundation guarantees that the structures built on top will wobble no matter how much is spent on models.
Info-Tech's recommended remedy is a federated data operating model that assigns accountability to domain experts while maintaining centralized standards, with the explicit goal of enabling AI readiness, improving data quality and unlocking cross-domain insight without sacrificing control. We find this the most actionable idea to come out of the event, precisely because it resolves the false choice between central control and local speed. The hub sets the standards and the spokes own delivery, which is the same federated pattern emerging from the broader CIO playbook this year. The vendors and analysts have converged on the answer. The 2.8-point gap shows how few enterprises have implemented it.
Scale or Fail Is Not a Slogan
Info-Tech's four-part Agentic IT framework, owning the AI mandate, picking the right bets, enforcing discipline, and proving and scaling value, reads as a checklist for the half of organizations that have momentum but no mandate. Each item is a corrective to a specific gap in the survey data. Owning the mandate addresses the missing board strategy. Enforcing discipline addresses the data governance shortfall. Proving value addresses the 42 percent who cannot yet show measurable cross-departmental impact. None of it is novel, and that is the point: the hard part is execution, not insight.
The phrase circulating through the CIO community this year, scale or fail, is a useful pressure test for any leader leaving an event like this energized. The enterprises that scale will be the ones that closed the governance and data gaps before they multiplied their agent count, not after. The ones that fail will be those who let bullish sentiment and rising budgets substitute for an approved strategy and a governed data foundation. Info-Tech's contribution is to make the gap measurable, which removes the excuse of not knowing. What remains is whether leaders will spend the next twelve months closing it or simply funding around it.
