Clinical AI is already part of daily care in many hospitals. The question I keep returning to is a simple one: do these tools work as well for every patient they touch? Most organizations cannot yet answer that with evidence.
Where hospital AI stands
In a 2024 American Hospital Association survey, 52.2% of U.S. hospitals used machine-learning-based predictive AI integrated with their EHR, 31.5% used generative AI, and 24.7% planned to within a year (Everson, Nong and Richwine, JAMA Network Open, 2025).
Adoption is moving faster than evaluation. That gap is where fairness questions go unanswered.
Evidence first, monitoring later
Continuous monitoring inside the EHR, with standards such as HL7 FHIR R4, is a direction the field is moving and a part of our 2027 roadmap. It is not something we offer today. Before any hospital can monitor a model responsibly, it needs a baseline: how does this model perform across age, sex, race, language and other patient groups in the data it has already produced?
What we do today
The Data Diversity and Fairness Auditor (DDFA) is a team-operated retrospective workflow that analyzes de-identified structured data and existing model outputs supplied under appropriate institutional approvals. It does not currently connect to live EHRs, ingest images or unstructured records, certify legal compliance, or provide point-of-care alerts.
If you are unsure whether a retrospective assessment fits, start with one question: which clinical model in this pathway produces structured outputs, and what subgroup-performance question does your current governance process leave unanswered?
Editor's note: this article was first published in August 2026. Our positioning was updated in October 2026 to reflect what Synod IntelliCare does today. We would rather tell you plainly what we can do now than promise what we have not yet built.