Trust Is the Constraint on CME's Next Product Cycle

Key Takeaways

• ACCME leadership treated trust as the condition that makes personalization, learner data, choice, and AI-supported CME useful at all.
• Credit administration was treated as time taken away from learning, not as a back-office inconvenience.
• AI-supported learning was welcomed only if it preserves reflection, teams, faculty, and conferences.

If clinicians stop trusting the educator or the activity, personalization, learner data, and choice become irrelevant. ACCME CEO Graham McMahon and board chair Annette Donawa put that condition on the next CME product cycle.

Personalization only works if the activity is trusted

In ACCME's Coffee with Graham conversation, McMahon and Donawa kept independence, content validity, fair decisions, and professional self-regulation as non-negotiable. The provider implication is the order of operations: data, personalization, choice, and AI do not rescue an activity clinicians do not trust.

That makes trust a product constraint, not a values paragraph. If a roadmap starts with adaptive recommendations, learner profiles, AI-supported screen learning, or tighter data exchange, the first test is whether the educator, the activity, and the decision process remain credible without those features. An earlier brief treated visible trust as the answer to a noisy information market; this conversation places trust in front of personalization and data, rather than treating it as a brand response to noise.

McMahon also treated credit administration as time stolen from learning and wanted CE data flowing to licensing bodies, certifying boards, and individual transcripts. Credit plumbing belongs in the learner experience, not after it.

They described accredited CE as a low-cost workforce performance engine that still arrives after the strategy is set. Donawa's Jefferson example, one institution's experience, shows joint accreditation helping interprofessional education gain senior buy-in. AI-supported learning can sit closer to care, but not if it becomes screen-only consumption or treats teams, faculty, and national conferences as optional. The work is to design those products inside the trust architecture, not on top of it.

What CME Providers Should Do Now

• For each AI, data, or personalization feature, name the trust condition it depends on: independence, validity, transparency, fair process, or faculty judgment.
• Identify the credit steps learners still perform manually, then decide which can move to board reporting, transcripts, or CMEpassport.
• Bring CE leaders into strategy conversations before the initiative is designed, especially for workforce performance, team development, or implementation change.
• Keep a faculty, team, or live-community path in AI-supported products so the screen is not the whole offering.

What CME teams should reconsider

Which AI, data, and personalization bets still hold if the activity must be trusted before it is tailored, if credit administration comes off the learner's plate, and if technology cannot hollow out reflection, teams, or conferences? That is the constraint ACCME put on the next product cycle.