In many cases, “AI” is used loosely. Sometimes it refers to machine learning systems used for imaging interpretation, risk scoring, planning, or documentation support. Other times, it may describe automated charting features or software that generates summaries based on clinician inputs. For a legal claim, the key question is not whether “AI” was mentioned in the chart, but whether the care team met the expected safety standard for the situation.
In Washington, hospitals and clinics routinely rely on electronic health records and digital workflows. That’s not inherently unsafe. The legal issue arises when software use, automated outputs, or computer-assisted recommendations are relied on without appropriate clinical verification, supervision, or follow-through. Even if a tool is capable of helping clinicians, a patient’s outcome still depends on whether the team acted reasonably and promptly.
When patients notice language such as “automated,” “computer-assisted,” “generated,” or references to decision-support workflows, it can feel alarming. Our approach is to treat those references as clues. We evaluate what the tool did, what data it used, how it was implemented, who reviewed the results, and what the clinical team did in response.


