Where AI Doesn't Belong
In 2025, Snow published ‘Read This Before You Automate Patient Content,' making the case that privacy must come first in patient storytelling, and that no amount of speed is worth risking patient trust. This is a mandate we will never lose sight of, no matter how sophisticated technology and AI become.
There are risks with using AI in creative, especially when it comes to patients. For example, fraud investigators have traced scam networks fabricating “before-and-after” testimonials for counterfeit GLP-1 products, complete with synthetic patient photos and doctor endorsements attributed to physicians who never gave them, according to Check Point Research. This poses a significant risk that patients make treatment decisions based on people who were never real.
Snow's 2025 patient survey of 300 patients reinforces why DTC marketers need to be deliberate in how AI is used. As one patient put it, “I just want to hear, and know that I'm hearing authentic stories from patients that I or my loved one can learn something from. Not something ‘manufactured' or heavily marketed.” Most patients feel the same way. Eighty-five percent said they are not open to patient stories that include AI-generated imagery, voices, or avatars, even when the underlying story is real. Patients rated real patient stories as 84% credible, while paid actors came in at just 19%, and AI-generated content only 15%.
Where AI Does Belong
We know there are practical places for AI in patient engagement, outside of replacing the patient's story. Voice cloning a translated patient video, with their explicit consent and transparent disclaimers, is the clearest one. A real patient story told in English can reach more patients and improve the viewing experience when it is translated, then delivered by the patient in their real voice, rather than dubbed over or with sub-titles. This preserves authenticity and makes their story more accessible.
The same goes for the work behind the patient experience. For example, Large Language Models (LLMs) excel at categorizing large amounts of information into pre-defined groups, helping researchers and strategists find trends in patient insights more quickly and at a higher volume than ever before. This same skill helps writers and producers understand gaps to uncover in their co-creation process with patients, personalizing the experience for Patient Ambassadors.
An operational opportunity is improving the accuracy and speed at which promotional materials are prepared for regulatory, legal, and medical reviews. By reviewing materials against approved sources and prior feedback, compliant content moves faster without cutting corners on what makes it credible. Across all of these, the model stays the same. Real patients and genuine relationships remain the foundation, with AI used on the back end to create richer, more personalized patient and client experiences.
What this means for DTC marketers
DTC marketing teams are under pressure to move faster and do more with less. AI can help with that by scaling patient stories to wider audiences, surfacing more insights faster, and shortening compliance timelines. What AI cannot do is replace the story, or the person, themselves. That is where cutting corners can impact brand reputation, patient trust, and the outcomes that trust drives. The marketing initiatives that succeed in this next phase of patient engagement will be the ones that know exactly where AI fits in the process.