Research

How Cardiologists Are Using Generative AI

Exploratory research among healthcare professionals at ESC Congress 2026.

Published 22 September 2026 by Theralexis

healthcare professionals: 27
cardiologists: 24
countries: 17

70%

reported that AI had influenced a treatment decision

19 of 27 respondents

How much it is used

AI is already part of medical information seeking

26 of 27used at least one AI system weekly or more for medical information

Used daily or weekly for medical information Base: 27 respondents

ChatGPT: 78%21 of 27
Google Gemini: 48%13 of 27
Claude: 33%9 of 27
Perplexity: 7%2 of 27
Microsoft Copilot: 0%0 of 27

ChatGPT was the most widely used system. OpenEvidence was the most frequently mentioned additional platform.

What it has influenced

Respondents say AI is influencing clinical thinking

Has information from an AI system ever influenced any of the following in your professional clinical practice? Base: 27 respondents

How I understood a disease or treatment: 93%25 of 27
How I approached a diagnosis: 85%23 of 27
Which treatment options I considered: 85%23 of 27
A treatment decision: 70%19 of 27

Each item was asked separately.

What it is used for

AI is being used to compare medicines

Have you ever used generative AI to compare …? Base: 22 respondents

Safety of different medicines: 91%20 of 22
Efficacy of different medicines: 82%18 of 22
Treatment options: 82%18 of 22

Trust

How does AI compare with pharma on trust?

If the same medical question were answered by both generative AI and a pharmaceutical company, which answer would you generally trust more? Base: 22 respondents

Generative AI: 45%10 of 22
Pharmaceutical company: 23%5 of 22
About the same: 23%5 of 22
Neither: 9%2 of 22

How medicines are named

Most respondents ask about medicines by molecule, not brand

When asking an AI system about a specific medicine, how do you usually refer to it in your question or prompt? Base: 27 respondents

Molecule / generic name: 52%14 of 27
Either, about equally: 30%8 of 27
Brand / trade name: 11%3 of 27
Does not ask AI about medication: 7%2 of 27

Only 3 of 27 respondents said they usually ask AI about medicines by brand name.

How medicines are named in AI questions therefore matters for pharma.

About the study

About this research

Generative AI is becoming a new source of medical information for healthcare professionals. For pharmaceutical companies, that creates an important new question: how are HCPs using AI when they learn about diseases, treatments and medicines?

Theralexis conducted this exploratory survey among 27 healthcare professionals, including 24 cardiologists, at ESC Congress 2026. We asked about their use of generative AI for medical information, its influence on clinical thinking and treatment decisions, the use of AI to compare medicines, and how AI compares with pharmaceutical companies on trust.

The research helps us understand the growing role of AI in the HCP information journey — and why how brands, medicines and evidence are represented across AI systems increasingly matters for pharma.

Also in this series

How Dermatologists Are Using Generative AI — EADV Congress 2026

What this means

AI Representation is becoming part of pharma's go-to-market strategy.

If healthcare professionals are asking AI about diseases, treatment options, efficacy and safety, how AI describes a therapy becomes a commercial question.

Theralexis measures how AI systems represent therapies, brands, competitors, evidence and sources — and what to do about it. We identify the biggest opportunities for your brand to win and provide channel and content recommendations to take action.

See for yourself

See how AI represents your market.

We provide AI Visibility and Share of Voice benchmarks for free. See a real-market example below.

Real differentiation in pharma GEO goes beyond visibility. Schedule a meeting to explore the outcome-oriented metrics that matter for your business.

See an example Schedule a meeting