AI-Generated Texts Motivate Seniors to Stay Active

Study finds AI-authored messages effectively encourage physical activity, but perceptions of AI impact how they're received.

Apr. 10, 2026 at 4:38am

Got story updates? Submit your updates here. ›

A minimalist, neon-outlined illustration of a person walking briskly, conveying a sense of energy and movement through the use of vibrant, glowing lines against a dark background.AI-powered health messaging aims to motivate seniors to stay active, but perceptions of the technology can impact its effectiveness.Ann Arbor Today

A new study from the University of Michigan and Penn State University found that AI-generated text messages can effectively motivate middle-aged and older adults to be more physically active, but people's perceptions of AI impact how they evaluate the messages. The researchers asked 630 adults aged 40 and older to rate 80 AI-written texts designed to encourage movement, and found that only about 5% were seen as culturally insensitive and 6% had quality issues. However, those with more positive attitudes toward AI tended to flag more messages as potentially insensitive, even when they knew the texts were AI-generated.

Why it matters

As health programs increasingly look to leverage AI to drive large-scale behavior change, understanding how older adults perceive and respond to AI-authored content is crucial. This study provides important insights into the opportunities and limitations of using AI to deliver personalized physical activity interventions to middle-aged and older populations.

The details

The study, published in the Journals of Gerontology, had participants rate the cultural sensitivity and overall quality of 80 AI-generated text messages designed to encourage physical activity. Of nearly 50,000 total ratings, only about 5% were flagged as culturally insensitive and 6% had quality problems. However, participants who had more positive attitudes toward AI tended to rate more messages as potentially insensitive, even when they knew the texts were AI-generated. The researchers also found that messages focused on reducing sedentary time, rather than increasing activity, and those about preparing for exercise rather than performing it, received more low-quality ratings.

  • The study was published on April 10, 2026.

The players

Allyson Tabaczynski

A postdoctoral research fellow at the University of Michigan School of Kinesiology and the lead author of the study.

University of Michigan

The institution where the lead author is based and where the study was conducted.

Penn State University

The institution that collaborated with the University of Michigan on the study.

Got photos? Submit your photos here. ›

What they’re saying

“Initially, I thought this was a little counterintuitive. If you have a more positive attitude toward AI, you might also just have more general knowledge of some of the biases or limitations that AI can have in its output or in its training data.”

— Allyson Tabaczynski, Postdoctoral research fellow

“If someone is receiving a health intervention that uses AI, their perceptions of AI are going to impact how they're evaluating or responding to that intervention. So it's something that researchers and interventionists have to take into account as they're designing their interventions with this technology.”

— Allyson Tabaczynski, Postdoctoral research fellow

What’s next

The researchers plan to continue exploring how to optimize AI-generated content to effectively motivate physical activity in older adults, while addressing potential biases and limitations of the technology.

The takeaway

This study highlights the promise of using AI to deliver personalized physical activity interventions at scale, but also underscores the importance of understanding how older adults perceive and respond to AI-authored content. As health programs increasingly leverage AI, accounting for users' attitudes toward the technology will be crucial for driving meaningful behavior change.