Opinion: Don’t throw your data away on AI podcasts 

By Published On: January 7, 2026Last Updated: January 14, 2026
Opinion: Don’t throw your data away on AI podcasts 

By EatMoreFruit Communications

There’s a trend of people creating AI podcasts – feeding data, journal articles, reports into AI tools and creating podcasts with avatar hosts.

I don’t think it’s a good idea.

On a scale of cool content to scary deepfakes, where do you think this falls?

One end says it’s a nifty thing to show in an office and say “look what I did in an hour with this data.” The other is a harmful threat to democracy where fake audio and videos are designed to spread misinformation and undermine authentic experts.

As marketers, we love gadgets. We’re always looking for new innovations, new ways of doing things.

AI has been talked about in PR  for a long time, so much so that a brilliant training course I did recently was called “AI, Always F&cking AI”

So I do understand that these AIcasts feel a lot more exciting than another evidence infographic.

But where do you draw the line between that, misinformation, and undermining the authenticity of your key opinion leaders?

They often use an avatar to not just give an overview of your data, but build in opinions as well. And to whom do those opinions really belong?

Brands spend a long time building authentic relationships with policymakers, surgeons, PhDs. People that have invested years to become experts in their fields.

And people aren’t sick of experts – they can’t be in healthcare.

And AI engines, where people are increasingly searching for health advice, certainly aren’t underestimating the importance of having a credible author biography on a piece of content.

The Problem: They’re Self-Serving

The problem I have with these AI podcasts is that they’re ultimately self-serving.

You are turning your own data into a 20-minute podcast with an AI avatar – and are you even going to listen to that? I’m not sure you are.

That’s not to say they’re all bad.

OrthoBullets is a great example of an AI generative podcast done right.

You can totally see a surgeon from the British Orthopaedic Training Association (BOTA) listening to these on repeat, or even a consultant  wanting to recap a specific and detailed topic.

It’s factual, it’s concise, it’s best practice clinical guidance.

They don’t need opinion in these episodes. They’re a resource to listen to for a purpose. It’s not necessarily for entertainment, and listeners might not be interested in every single episode.

In short, these episodes are for a specific state of mind, but that’s not necessarily in an emotive state.

Podcasts on the whole, are inherently quite emotive things. And that’s because they’re so personal and intimate.

How Often Has Someone Said…

How often has someone said to you “oh, you should listen to this podcast, I think you’d really like it”? And you go “yeah, sure.”

Maybe you even go as far as to download it and save it. I did that this week with In Our Time, which was recommended to me by EatMoreFruit CEO, Paul Jarman.

But I’m probably never going to listen to it, even though I’ve got the best of intentions and I like the sound of it.

Really, it’s not the content or the topic that makes people love podcasts – it’s the hosts and the characters.

I love Lizzie Bassett on What Went Wrong. Libi loves Jessie Ware and her mum on Table Manners. Rachel loves Gi Fletcher on Happy Mum, Happy Baby.

We’ve all got these different podcasts and niches and topics that we engage with. We’ve all listened to episodes of these podcasts that weren’t on topics we were interested in at all.

But we listen to new ones because of who the hosts are, and the relationship that we’ve built with them.

How Do People Find Podcasts, and What Does That Mean for Visibility?

Friend recommendations do make people listen to podcasts, but that tends to be for newer listeners.

People that have been listening to podcasts for a while tend to find them by themselves and discover them through topics, then stay for the format and the host.

Discovery trends

  • 40 per cent of listeners discover new podcasts via in-app search
  • 30 per cent discover new podcasts through recommendations from friends and family
  • 19 per cent find new podcasts through social media platforms.
  • YouTube is the leading platform for discovering podcasts.
  • 10 per cent of listeners discover podcasts through influencer recommendations.

But these also show that the algorithm is strong.

That’s how people tend to find them. But they also search by category or topic – they might search “history podcast” or “comedy” or “news podcast,” which might bring them in as new listeners.

People are willing to give it a go, but they’ll switch off quite quickly if they don’t connect with the hosts and the format.

This is where OrthoBullets get it right again. Same podcast, totally different episode format: CoinFlips and controversies.

Each episode focused on a grey zone surgical topic, hosted by a real surgical expert in the field and joined by colleagues to discuss.

Why Is the Influence So Strong?

It’s all to do with that parasocial relationship that drives long-term listening engagement. It’s inherently intimate – it’s in your ears. And just like in daily life, who you choose to listen to says a lot about you.

It has to speak directly to how you see the world, or want to. Which makes it more intimate than what you watch.

It’s not a passive activity. The behavior of listening becomes an active routine for many people.

Routine habits

  • 49 per cent of listeners are cleaning
  • 42 per cent are on their commute
  • 29 per cent are exercising
  • 25 per cent are out walking
  • 15 per cent of listeners are working

That  routine is important.

When people build a connection to a trusted, predictable format that gives way for more introductions to new topics, new ideas, new brands because they like that host, that format, or that activity.

With AI generated podcasts, there is no relationship to build with that avatar. And people probably aren’t looking for your data specifically.

So What Does That Mean for Healthcare Brands?

First of all, please don’t start your own podcast unless you really have a brand new concept and can keep it going, and have a great host for it.

Most importantly to me, don’t throw away your new data in a one-episode podcast that even your colleagues won’t listen to all the way through.

What to do instead?

Get personal with the pitch. Interview opportunities are a really powerful resource for a podcast host, but it has to fit their format.

Clinicians, patients, or academics might be looking for that peer-to-peer podcast.

There are a huge number of specialist trade podcasts out there –  like Anaesthesia on Air from the Royal College of Anaesthetists or Dr. Emily Schehlein from Ophthalmology Journal are both examples where organisations are doing great work to put these podcasts out, with great interviews made available and accessible to healthcare professionals.

But don’t overlook the general interest and news angles.

One of my favorite examples of this is when we got client data featured on The Week Unwrapped. They were talking about psychedelics because of the evidence.

In this category, it wasn’t just healthcare professionals and psychiatrists we wanted to reach with this data, but we were trying to bring legitimacy to a new category of using psychedelic medication to treat addiction and mental health conditions.

Having that story featured on a general interest news podcast was a real way of bringing category legitimacy to life.

You don’t have to listen to me…

Healthcare has all those ingredients. It has passionate people. It has new data. It has a hundred ways of being able to connect different stories to different formats.

But next time you have clinical data or new evidence, and there is a risk people won’t read it…don’t just feed it to NanoBanana. Instead ask a PR team to get it in people’s ears in the right way.

They’ll get it to engaged listeners who’ve already chosen to trust a host, who are already in that routine. They’ll turn your data into a story that fits a format people have actually subscribed to.

That’s how you build influence – not by manufacturing a relationship with an AI avatar, but by earning a place in a relationship that already exists.

Online NHS hospital to focus on menopause, prostate and eye conditions
Research to harness wearables data to understand knee surgery impact