AI is rarely the only “cause.” More often, it’s part of a chain: an automated triage tool, risk scoring, imaging support software, lab workflow automation, or documentation assistance. The legal issue is usually whether the care team used that information responsibly.
In Bloomington, we commonly see patterns like:
- Symptoms first, diagnosis later: A patient is evaluated quickly—sometimes more than once—before the correct condition is recognized.
- Imaging or lab results not acted on promptly: Reports may arrive, but follow-up decisions (or communication) lag.
- Risk scores used as shortcuts: Tools may flag “low probability” conditions, and clinicians may not fully verify alternatives.
- Documentation gaps across visits: Notes from one facility don’t clearly carry into the next appointment, delaying escalation.
When that chain results in harm, the question becomes: what should have happened at each step, and whether the deviation from accepted practice affected outcomes.


