Health Systems Bet on Enterprise-Wide AI Strategy Over Piecemeal Pilots

Piecemeal AI deployment is starting to look like a dead end for hospital systems chasing return on investment.

That is the emerging consensus among health system leaders grappling with digital transformation, and it forms the core argument from Dr. Deepti Pandita, vice president of clinical informatics and chief medical informatics and AI officer at UCI Health/UC Irvine. Speaking ahead of her appearance at the HIMSS AI in Healthcare Forum in San Diego next month, Pandita makes a case that should interest anyone tracking enterprise AI spending: use-case-based, department-specific rollouts are giving way to enterprise-wide strategies that unify technology, operations and governance.

The logic is straightforward. Healthcare organizations have grown too interconnected for isolated fixes to work in isolation. "A change made in one department often affects many others," Pandita said. AI tools, digital workflows, patient engagement platforms, analytics and EHR modernization efforts all cross traditional departmental lines. Let each department chart its own course, and the result is fragmented experiences, duplicated spending, inconsistent governance and systems that cannot scale.

That is a costly failure mode. Health systems investing in AI without enterprise coordination risk paying twice for overlapping capabilities while gaining none of the leverage that comes from a unified platform. The pattern echoes broader concerns in health IT circles about foundational gaps undermining AI ambitions, gaps that other CIOs have flagged as a prerequisite issue that has to be solved before scaling anything.

Why Alignment, Not Technology, Is the Real Bottleneck

Pandita's central claim cuts against a common assumption in enterprise tech deployment: that the hard part is the software.

"The hard work is not usually the technology itself, it's creating alignment," she said. Getting there requires pulling clinical leaders, operations, IT, finance, compliance, quality, nursing, physicians and executive stakeholders into a shared vision with shared success metrics. That is an organizational design problem, not an engineering one.

Governance, trust and transparency underpin that alignment. Just as important is a willingness to make decisions that serve the enterprise even when they are not optimal for any single department. That is a hard sell in organizations where departmental budgets and incentives are often siloed by design. The health systems making real progress, Pandita said, treat digital transformation as an organizational capability rather than a portfolio of discrete projects.

The investment implications of that framing are significant. Systems making headway put as much capital and leadership attention into change management and culture as they do into the underlying technology stack. In many cases, culture, not software, becomes the true constraint on scaling. That is a meaningful signal for vendors and health system executives alike: technology procurement decisions are necessary but not sufficient conditions for transformation success.

This mirrors what other health systems have reported publicly. Mass General Brigham's experience embedding AI across its enterprise has been cited as a valuable reference point precisely because it required coordinated governance rather than isolated departmental buy-in. The pattern recurring across these case studies suggests a market-wide realization that AI ROI is gated less by model performance and more by organizational readiness.

Pandita frames the stakes bluntly. The organizations that ultimately win in healthcare digital transformation, she argues, may not be those that deploy the most technology. They may be the ones best equipped to build trust, alignment and organizational readiness around whatever technology they do deploy. Those capabilities, she said, determine whether an innovation stays a "promising pilot" or becomes genuine enterprise transformation.

That distinction matters for anyone evaluating health IT vendors or provider organizations as investment targets. A hospital system with an impressive pilot program and a press release is not the same as a system with the governance infrastructure to scale that pilot across dozens of departments and thousands of clinicians. The gap between the two is where most digital transformation budgets quietly disappear.

The Bottom Line

The enterprise-wide framing Pandita describes is not a novel technology thesis; it is an organizational one, and that distinction matters for how outside observers should assess health system AI initiatives.

Investors and industry watchers evaluating health IT vendors should pay close attention to how prospective health system customers structure their governance before assuming a deal signals durable technology adoption. A hospital that buys AI tools without enterprise-level alignment, shared metrics and cross-departmental buy-in is a higher-risk customer: pilot fatigue, duplicated procurement and stalled rollouts are the likely outcomes, not smooth scaling.

For vendors themselves, the implication is equally direct. Selling point solutions to individual departments may generate near-term revenue, but Pandita's framework suggests that the more durable, higher-value relationships will go to vendors who can support enterprise-wide governance structures, not just single-use-case deployments. That favors platform providers and systems integrators over narrow point-solution vendors in the long run, though the latter will continue to find buyers among departments moving faster than their enterprise counterparts.

There is no hard ROI figure attached to this shift yet, and Pandita's remarks are directional rather than quantitative. But the qualitative signal is consistent with what other health system leaders have said publicly about foundational readiness preceding AI scale. Anyone underwriting healthcare AI adoption as a purely technological story is missing half the equation. The other half, governance, trust and organizational culture, is harder to price but appears to be the real determinant of whether these investments pay off.

Pandita will discuss the topic in a fireside session, "An Enterprise-Wide Approach to Digital Transformation," at the HIMSS AI in Healthcare Forum in San Diego on October 22.