Project · Internal council tool

AI Curriculum Mapping In use

Every major organization now publishes its own AI-competency framework for medicine. Our comparison platform puts them side by side in one unified structure — so a school can see consensus, differences, and gaps, and build its own framework on the evidence.

The problem

AAMC learner and faculty competencies, peer-institution frameworks, published academic frameworks, and industry guidance all slice "AI competence" differently. A school choosing what to teach faces incommensurable documents: different domains, different granularity, different vocabularies.

The tool

The AI Competency Framework Comparison Tool synthesizes eight-plus source frameworks — including the AAMC learner and faculty competencies, published frameworks from peer institutions in the US and internationally, DATA-MD, and industry framings — into a unified superset framework of 7 domains and ~47 topics across 3 depth tiers, then offers four working views:

• Synthesis dashboard — where the frameworks agree: consensus competencies at a glance;
• Domain drill-down — side-by-side comparison of how each framework treats a given topic, with the original competency text always quoted verbatim;
• Framework profiles — the structure and emphasis of each source framework on its own terms;
• Framework builder — an editable workspace where an institution drafts its own framework, borrowing verbatim language from any source, with export to Excel, Word, and JSON for committee work.

Method rigor

• Coverage judgments are explicit and three-valued (full / partial / none), with inferred mappings flagged as such.
• Framework text is never paraphrased in the mapping — always verbatim quotation.
• Mappings are verified by cross-vendor multi-model consensus: three frontier AI models from different vendors independently review each mapping; two-of-three agreement is required to apply a change, and disagreements are logged for human review rather than auto-applied.

What it feeds

The unified framework is the backbone for our curriculum integration work — every existing session and course maps onto it, making gaps and priorities visible — and for the institutional competency set being developed alongside the AI guidelines. The tool itself is currently an internal, access-controlled resource for the AI in Medical Education Council; we're glad to demo it for collaborating institutions.