The invisible influencer: the new information layer between doctors and scientific evidence
By Dimithri Wignarajah, Michelle Collins and Jill Noble

Knowing your customer is no longer enough. We also need to understand the information ecosystem shaping their knowledge and decisions.

Scientific engagement has always been an ecosystem. Healthcare professionals (HCPs) seek information, while key and digital opinion leaders, societies, publishers and peers help shape scientific understanding. That ecosystem has never been linear and prescribing decisions are rarely driven by a single source.

Now, AI is becoming another influence within the information ecosystem.

AI consumes information, interprets it and increasingly surfaces (often invisibly) when HCPs and patients search for answers. We know large language models (LLMs) are influencing scientific learning: 51% of oncologists use generative AI to find background information and 49% use it to compare sources and viewpoints.

An HCP can query this ‘super intelligence’ intermediary and receive a synthesised answer without consciously choosing a publisher, journal or website source.

This does not mean AI is replacing scientific societies, publishers, peer networks, congresses or direct engagement. Trusted scientific voices and established channels remain essential. But AI’s influence is increasingly important within the context of the wider information and learning journey.

It changes our thinking from ‘How do we reach our audience?’ to ‘Who and what is shaping their thinking?’

Understanding the audience remains the foundation of scientific engagement. First-party data tells us what audiences need, do and prefer. But now a second layer of insight is needed, an understanding of the information ecosystem’s views in shaping what HCPs discover and the answers they receive.

We need to identify where AI is likely to influence information-seeking behaviour, then ensure the right evidence, content and engagement strategy exists at those key moments.

The most effective organisations will combine audience insight with ecosystem insight. GEO audits and discoverability analysis can help reveal what is influencing the answers audiences receive. They can identify the types of questions being asked, the scientific sources being cited and the evidence that surfaces.

Those insights inform content and omnichannel engagement strategy, ensuring scientific content is activated where audiences engage and where answers are increasingly being formed.

AI adds another step to the planning equation. Discoverability, credibility and influence now need to sit alongside audience behaviour.

We need to ask ‘Where will our audience engage?’ and ‘Where does our science need to exist to influence what they discover?’

This means balancing targeted HCP environments with authoritative, discoverable sources such as scientific publishers, YouTube and professional social platforms that help evidence travel through the wider information ecosystem. Permissioned direct channels then play a distinct role: one-to-one services such as WhatsApp Business can deliver timely congress updates, new-data alerts and follow-up resources to opted-in HCPs, reinforcing priority evidence at the moment it is most relevant.

In a recent GEO audit conducted for a client, 55% of earned citations of their brand came from only eight sources. It highlights how a relatively small number of sources can exert huge influence on understanding and perception.

Audience needs, priority questions and content strategy remain central to engagement planning. Scientific content increasingly serves two audiences: the people consuming it and the systems retrieving, interpreting and surfacing it.

Modular content already helps teams adapt and reuse scientific content across audiences and platforms. AI adds another dimension: how the core knowledge is structured so it can be easily discovered, interpreted and accurately reused by machines as well as people.

That means designing content for both engagement and discoverability. Scientific knowledge needs to be credible, referenced and written so it can be useful when questions are asked.

‘Share of Answer’ provides a useful way to think about this. When priority scientific questions are being answered, how visible, accurate and useful is our evidence?

AI-mediated discovery often escapes traditional metrics. GEO/AEO audits reveal what answer engines cite, which sources and competitors shape responses, and where evidence gaps exist. These insights help medical communications planners prioritise questions, content, placement and measures of influence while ensuring science is credible and accurately represented at discovery.

Omnichannel remains the foundation, but it now carries new planning requirements: understanding not only the audience, but also the ecosystem shaping the answers they receive.

AI will not replace trusted scientific voices or established channels, but it does add a new, somewhat opaque layer of influence to how scientific knowledge travels.

For strategic medical communications it creates a new opportunity: to address how people learn, we need to rethink how content is structured and targeted for the people and tools guiding them, so when the questions that matter are asked, the right evidence is part of the answer.