Lessons From MAPS 2026 On Turning AI Into Action

Sean Markwardt, PhD | SVP, Growth Optimization, Medical Communications

If you walked the Medical Affairs Professional Society (MAPS) Americas 2026 Annual Meeting exhibit floor in Denver, you couldn't miss it. AI-powered insights, data integration platforms, next-gen analytics, and real-time intelligence ecosystems: the buzzwords arrived en masse.

It’s clear that Medical Affairs is entranced by the AI era, even if not yet fully committed. This makes what we heard repeatedly in sessions, workshops, and conversations worth pausing on. Read on to learn more. 

Curiosity Shifted to Governance, but a Harder Question Got Skipped

AI was in discovery mode at last year's MAPS 2025 Annual Meeting in New Orleans. The discussion focused on what was possible, what was working, and what was neither. A year later, the vocabulary has matured into governance frameworks, scaled adoption, and debates about whether the field is moving too fast. The progression was palpable.

Across both years, the field celebrated or debated the tools while a more fundamental question sat largely unanswered: how does any of this produce the actions required to change decisions?

Data circulating in Denver made the gap concrete. The vast majority of Medical Affairs organizations are measuring activity tracking. Fewer are assessing HCP knowledge and beliefs. Fewer still are measuring actual HCP behavior change, and almost none can cite specific clinical impact. The field is generating data at the wide end of the funnel and losing momentum before it becomes action. That’s not an AI problem, it’s an organizational one.

Part of what's driving this is a structural limitation in how insights have traditionally been gathered. The typical loop — event occurs, insights are obtained, strategy is updated, teams prepare for the next event — sounds functional until you look at where it breaks down. The loop exists on paper; in practice, it is a series of disconnected snapshots.

The Conversation That Said it Plainly

At our booth in Denver, a field medical leader told us her team was capturing strong insights, only to die inside a spreadsheet.

After a few minutes, the real problem came into focus. It wasn't the data, and it wasn't the tool. What was missing was the communications and change management architecture to move insights from collection to action. The data and AI had done their job. The human infrastructure hadn't been built.

Three Layers, Not One

What no AI platform can fix on its own are the human and organizational layers. The human layer is a scientific and strategic interpretation. Knowing which signal matters, in which therapeutic context, for which stakeholder, and why it warrants action right now. That judgment does not ship with a software license.

Ask ten Medical Affairs professionals to define what clears the threshold as a true "insight", and you’ll likely get ten different answers. That gap alone creates friction at every handoff.  

The organizational layer is communications design and change management. In fast-moving moments — data releases, congresses, launches, competitive shifts — the teams that respond well are those that have built the infrastructure to interpret emerging signals in near real time and translate them into aligned internal action. Speed of intelligence means nothing if the organization isn't built to absorb it.

"End-to-end AI insight integration" and "data intelligence platforms" solve for one layer and oversell themselves as the only missing puzzle piece in your organization. Our field deserves a more honest conversation than that.

The Scorecard That Actually Reflects Progress

Medical Affairs teams might spend the next twelve months implementing what Denver shouted about. But ask your organization now: are the insights your tools generate actually changing decisions, closing care gaps, or accelerating the delivery of a scientific narrative to someone who can use it? That’s the scorecard that will separate organizations that got it right from those that didn't.

Without the human alignment and organizational infrastructure to interpret, communicate, and activate, fast intelligence is just expensive noise.