Can ChatGPT help news publishers build trust, community?

By Jodie Hopperton


Los Angeles, California, United States


Hi there.

Since I ran the INMA study tour in Silicon Valley a couple of weeks ago, my mind has been buzzing thinking about how AI will affect our products and our industry. The rate of change is so rapid that since a debrief I gave last week, OpenAI has made consequential announcements, dramatically bringing down costs and making it easier for anyone to build and monetise through a kind of ChatGPT app store. 

Product people I am speaking to know fundamental shifts are coming, but they are struggling to work out where to place their bets. That’s what I want to talk about today. What are the issues, the opportunities, and where we should place our bets. 

I love discussing this and want to get as many viewpoints to help shape my thinking and reporting back to this community, so if you have something to say, please reach out to me at

Thanks, Jodie

A step forward for the news industry?

We have an overall hypothesis in the news media industry that as people age up and become civically engaged — buy a house, send their kids to school, become more interested in politics — they have a higher propensity to subscribe to our products. This comes partly because we have yet to figure out a way to monetise effectively off-platform. 

But were already disproving that hypothesis. The majority of digital subscribers are 45+. There are very few (any?) news organisations that are attracting younger paying subscribers. Not Millennials and certainly not Gen Z.

But look at social platforms, any of them. The top accounts are always individuals. Something Richard Gingras wrote about recently and then discussed with us on the INMA Silicon Valley Study tour is trust.

He argued that there is not a dearth of trust overall — everyone trusts someone. This really struck a chord with me. People trust people. They don’t always trust brands. And that makes sense on a human level: You see a face, you get to know someone. It’s a mirror, something similar we humans can connect to. And that’s why social has done well. It’s also why Substack has done so well. 

And therein lies a challenge for us as news brands: How can you effectively build that connection? We talk about engagement. Maybe we need to start talking about connection. 

I am well aware that a brand, an institution, means processes. For us it means rigor. It means separating fact from fiction. Chasing the truth, providing context to complex situations. People may understand this, but we need to find a way to connect that back to the human level. 

So how do we make a brand more human? 

The obvious thing is to show the faces behind the news. But that also brings challenges as personalities can come and go. Those with large followings are exactly those that do well on Substack — on their own. Every publication deals with this, and it’s nothing new. 

Another way is using audio. Voice has been shown in numerous studies to provide greater connection to people. Good quality audio builds trust. Schibsted spent a lot of time thinking about what their synthetic voice should sound like. More about that in our recent report here.  

And a new way to connect, with the help of GenAI, is chat. As we’ve already seen with ChatGPT and other GenAI products, we’re moving from searching for a few keywords to asking questions. And the answers we are getting back are detailed, as if a human had put them together. We can further interrogate these responses, build on them further. This is chat. And chat is human-like. It builds connection. As a whole, this may move us a step closer to trust.

News media companies can use chat to build trusted connections.
News media companies can use chat to build trusted connections.

On the study tour, we met a company that, I think, solves a number of the issues above. It’s a chat application that can either be placed as a chatbot on our Web site. Or it can be used on any chat platform. Yes, pretty much any: WhatsApp, Teleram, Facebook Messenger to name a few.

“But we have broadcast ability on WhatsApp now!” I hear you say. 

Yes we do. But that is a one-way broadcast communication. We cannot converse.

The reason I got excited is actually not just about the chat. It’s that they can provide off-platform analytics for these tools. Advertising and sponsorship can be bought into the chat. And it could also be used as a subscriber-only benefit. On the surface, it seems to be a product that we can use off-platform, can get metrics for, monetise, and build into our subscription products to increase engagement. This is the kind of thing that could be a game changer. 

This is the kind of gem we love surfacing on study tours. It’s actually been used predominantly in sports and commerce to date. There are trials with media at the moment, of which I will keep you posted to see if the promise of the platform lives up to my expectations. 😊

Date for the diary: Product Advisory Council “Ask Me Anything” on Wednesday, November 29

You’ve heard me talk about them. You may know how much the INMA Product Initiative work is based on this same group of product leaders. Now you get to ask them questions. 

What do they think is the key to product success? What’s one thing they wouldn’t compromise on when it comes to org structure? What are they most proud of this year? And what would they do differently? 

These are just some of the questions you may want to ask four members of our council:

  • Katharine Bailey, global head of product and design, Conde Nast (USA).

  • Riske Betten, B2C/product director, Mediahuis (Netherlands).

  • Kara Chiles, SVP of consumer products, Gannett (USA).

  • Julian Delany, CTO data and digital at News Corp (Australia).

Sign up here and feel free to send any questions ahead of time to me at

The product manager’s AI dilemma 

There are so many new AI tools on the market that are tangential to product. They can help with content creation and distribution as well as specific product features, such as summaries and narrated articles. And there are likely going to be a lot more to come.

But where do you start? How do you know what is worth building into the internal processes? How do you prioritise which to experiment with?

It’s hard to focus and to understand what is actually going to make an impact. In fact, most product leaders I speak to are trying to understand where they should make their bets. 

We can be pretty confident AI will provide cost savings and help us improve products, but there is an upfront cost — and it’s hard to put numbers against unknown outcomes. 

Something that struck home recently:

When asked about making bets on innovation during the Silicon Valley Study Tour, Justin Ferrell at Stanford’s replied that you can’t make decisions based on data from the past unless the future will look like the past. It’s simply not rational. And yet, in this time of exceptional technological change, that is what we are trying to do. 

Some media companies are buying enterprise licenses to ChatGPT or similar (the former of which is actually quite hard to do as they have only two sales people and currently no automated process). This is an excellent way to get people using it. Because once we have teams using it and see users interacting with the fruits on these trials, we can start to understand where the benefits are. 

Which brings me to another point that Justin made when talking about AI. He used an analogy of a bike: You can learn all about a bike, how the wheels and gears work, the best angles and height for handlebars, etc. But until you actually get on it, you don’t know how to ride it. And that is effectively what we are dealing with at this moment in time. 

For many leaders, this will be uncomfortable. Some control will need to be ceded. There will need to be AI guidelines drawn up that everyone adheres to, and you’ll have to assume that people will act as grown-ups.  Here is a good example of an approach from Ringier in Switzerland. 

It’s an uncomfortable moment in time. AI is rapidly changing the game, but we don’t yet know how. There will be a lot of experiments, which means some degree of failure and hopefully some successes. But things are changing — quickly — and doing nothing may mean that you can’t catch up. So now it’s all got to be about educated guesses, empowering people, and learning. Just keep in mind the end goal: What benefits are you bringing your users?

About this newsletter 

Today’s newsletter is written by Jodie Hopperton, based in Los Angeles and lead for the INMA Product Initiative. Jodie will share research, case studies, and thought leadership on the topic of global news media product.

This newsletter is a public face of the Product Initiative by INMA, outlined here. E-mail Jodie at with thoughts, suggestions, and questions. Sign up to our Slack channel.

About Jodie Hopperton

By continuing to browse or by clicking “ACCEPT,” you agree to the storing of cookies on your device to enhance your site experience. To learn more about how we use cookies, please see our privacy policy.