Earlier today I presented briefly to the Council of Deans of Health, on the topic of AI use in Fitness to Practice processes. The title of the presentation was, AI in fitness to practice processes: Navigating complexity with purpose. This was a follow-up to an invited blog post published a few weeks ago on the same topic. I wanted to use this opportunity to explore this idea in more detail, especially around the implications on different stakeholders.
Generative AI tools like ChatGPT and Claude are already being used throughout complex organisational processes, although often episodically, inconsistently, and without clear frameworks. I’m going to explore this idea through fitness to practice processes in nursing education, which I believe offers a compelling example of why systematic approaches matter. These types of high-impact processes are going to increasingly come under the spotlight, where AI use can both provide enormous value while also adding significant risk if poorly implemented.
FtP processes involve multiple stakeholders: students facing career-defining decisions, staff maintaining professional standards, clinical educators documenting concerns, and panels making complex judgements about competence. Each group might reasonably use AI tools, yet without institutional frameworks, this use happens in isolation, creating risks, missed opportunities, and inconsistent approaches that undermine process integrity.
The opportunities are substantial. AI can address equity concerns, help professionals structure comprehensive documentation without the overwhelming administrative burden often associated with high-impact processes, and prompt systematic consideration of multiple perspectives. But realising these opportunities requires more than individuals experimenting independently.
Organisations need decision frameworks that guide purposeful use at each stage of complex processes. For FtP, this means asking: how can AI strengthen the evidence base while preserving human judgement? How can AI support deeper thinking rather than replace the cognitive work we’re trying to assess? How can AI enhance transparency of reasoning rather than obscure whose professional judgement underpins decisions?
The fundamental organisational choice isn’t between allowing or prohibiting AI; it’s between systematic integration and fragmented, underground use. Control-focused responses risk driving use into hidden corners where learning can’t happen and quality can’t be assured. Systematic integration creates explicit frameworks that help everyone use these tools purposefully while maintaining focus on core organisational purposes.
This requires governance that builds organisational capacity to evaluate AI outputs critically, not just technical skills. It needs documentation and transparency systems that reflect how work actually happens, and frameworks that provide clarity while formal policies develop—because organisational silence isn’t neutral.
FtP processes are just one example. Research supervision, clinical decision-making, case management, quality assurance—any complex process involving multiple stakeholders and professional judgement faces similar questions about whether AI integration strengthens or undermines core purposes.
Systematic approaches move organisations from reactive responses to deliberate integration, from hidden use to transparent frameworks, from individual experimentation to organisational learning. The question isn’t whether AI belongs in complex organisational processes, but whether organisations will lead their integration thoughtfully or have it happen to them in ad hoc, isolated, and discrete activities.