Black-and-white portrait of Abbie Pratt, healthcare AI and AI governance executive, with a quote warning that a clinical model can perform as designed yet fail patients who were not adequately considered in its design.
Portrait of Abbie Pratt, healthcare AI and AI governance executive, with a quote emphasizing that the real-world performance of clinical AI matters more than how impressive it appeared before deployment.
Portrait of Abbie Pratt, healthcare AI and AI governance executive, with a quote emphasizing that the real-world performance of clinical AI matters more than how impressive it appeared before deployment.

What I’m Looking At in Healthcare AI

There are a few questions I keep coming back to as AI becomes part of everyday healthcare.

AI and Clinical Decisions

What changes when a clinician begins using AI as part of a decision? AI can provide information, identify patterns and support judgment, but over time it may also change how people make decisions without it. I’m particularly interested in what that means for clinical judgment and the expertise clinicians need to retain.

Does Advanced AI Actually Mean Better Care?

A model can produce an impressive result. I want to know what happens after that. Does it help someone make a better decision? Does it improve the way care is delivered? Does the patient actually benefit?

There can be quite a difference between a good technical result and a meaningful healthcare outcome.

When Is AI Ready to Be Used?

Not every healthcare AI system needs the same level of technical capability. An administrative tool and an AI system influencing a clinical decision carry very different consequences.

The closer AI gets to patient care, the more I want to understand what it was tested on, how reliable the results are, where the uncertainty is, and whether the evidence supports the way it is actually going to be used.

After AI Goes Live

Deployment isn't the end of the story. The data may change. The model may be updated. Clinical workflows change. People also find their own ways of using technology once they have lived with it for a while.

Healthcare organizations need to know whether an AI system is still doing what they originally expected it to do, and when something has changed enough to deserve another look.

How AI Changes Healthcare Work

I'm also watching what AI does to the workforce itself.

Some tasks may disappear. Others may move to someone else. New responsibilities will emerge. And as people rely more heavily on AI, some skills may become more important while others are used less often.

For healthcare, that raises practical questions about workload, clinical expertise, accountability and how we prepare people to work alongside these systems.