AI Is Not the Strategy: Knowing Where to Deploy It Defines Value

Key Highlights

• Artificial intelligence functions as an instrument within a clinical strategy rather than as a strategy in itself.
• The decisive question for health systems is where in the care pathway a given tool produces measurable benefit.
• Two clinical contributors, a hematology professor at Cairo University and UC Irvine's chief medical informatics and AI officer, frame deployment judgment as the central discipline.

Context and Core Dynamics

The central argument advanced by Nelly Abulata, MD, PhD, MBA, professor of hematology in the Division of Clinical and Chemical Pathology at Kasr Al-Ainy Medical School and University Teaching Hospital, Cairo University, and Deepti Pandita, MD, chief medical informatics and AI officer at UC Irvine in Orange, California, is that tool selection is subordinate to strategic purpose.

Institutions frequently approach AI by acquiring a capability first and searching afterwards for a problem it might address; the argument reverses that sequence. A model, platform or decision-support product carries no intrinsic clinical value. Value emerges only when the intended task, the available data, the workflow into which the tool is inserted, and the outcome to be measured are specified in advance.

The contributors write from two complementary vantage points: an academic hematology practice in Egypt and an operational informatics leadership role in a United States academic health system. That pairing reflects a wider shift in health technology discourse, in which the relevant expertise is no longer exclusively computational but clinical, operational and organizational.

Caregiving Application and Social Impact

The same discipline translates directly to eldercare, rehabilitation and psychological support. Where a provider is weighing ambient monitoring for fall risk, conversational agents for cognitive stimulation, or documentation support for stretched nursing teams, the strategic question is not whether the tool is capable but whether the task, the population and the outcome have been defined. In long-term care that means mapping where a tool removes an actual burden, such as medication reconciliation, transfer documentation or night-time observation, before committing capital and staff time.

For psychological support services, the implication is a clear threshold for escalation to human clinicians and an explicit limit on what an automated system may decide. Evaluated this way, adoption becomes a governance exercise rather than a technology purchase, and providers able to state where a tool operates, who remains accountable, and what would count as failure are better placed to sustain deployment beyond the pilot stage.

Ethics and Compliance Note

Deployment must respect data protection, informed consent and clinical accountability. The disclosure record accompanying the commentary illustrates the expected standard: one contributor reports no relevant financial relationships, while the other reports board or advisory roles with EBSCO and Grid Health, speaking engagements with ACP and HIMSS, and income of $250 or more from HIMSS. Comparable transparency should govern procurement in caregiving, particularly where vendor relationships, resident data and vulnerable populations are concerned.