Beyond prompting: How Reuters is empowering its staff to harness the AI revolution
“In any moment of a big transformational change, you’re dealing with people.”
While much discussion around AI’s impact on the news industry has focused on the various AI tools available, Jane Barrett argues that not enough attention has been paid to the people living through this transformative moment.
This includes the crucial need to help them navigate how the technology is transforming their jobs – whether they like it or not.
“We all have our own feelings about it. Some of us are early adopters and we’re really keen on AI,” she said, speaking at our Paris AI Forum.
“But some of us have in our guts a real distaste for AI, thinking, ‘Oh my gosh, here’s another thing that’s going to eat our breakfast, lunch and dinner, we’re all going to be out of jobs.’ ”
As Head of AI Strategy at Reuters, Barrett is helping the global news organisation’s staff to adapt to seismic shifts that are transforming the industry and many of their jobs.
With a workforce of 3,500 people (2,500 in the newsroom and 1,000 in commercial, product and engineering), a wide variety of attitudes and levels of expertise regarding AI are bound to exist within the organisation.
“There is simply no one-size-fits-all approach you can take to rethinking operations. Those people need lots of different things,” Barrett said.
For leaders, this means that they need to be tuned into the whole range of mindsets and skills in their organisations – but also “to give the hard message that AI is here, and it’s for everyone.”
“This is no longer something that is done in a little team. AI is now across the piece,” she said.
Mandating AI use and ‘baseline’ literacy
To ensure widespread AI adoption in the organisation, Reuters is setting specific and fairly strict goals for internal AI use: the aim is for everyone in the company to use AI on a regular basis by the end of 2025.
Specifically, staff are expected to use AI – either external tools or on Reuters’s own Open Arena environment – 20 times a month. Barret can also monitor AI use via a dashboard.
“So this is now an edict. This is something everyone has to do,” she said.
That may sound drastic, but “you can’t change something if you can’t measure it. So measuring people’s usage and then helping them to become literate is really critical,” she said.
“I’m still getting comments like, ‘ChatGPT doesn’t understand what a quote is.’ Or, ‘Claude made up this source.’“
“And you’re like, ‘Yes… because that’s how generative AI works.’ ”
“So it’s about going back to the basics and helping people to understand what AI is – but, really crucially, what it isn’t; and what it can do – but, really crucially, what it can’t do.”
AI skills as a career multiplier
In terms of concrete skills, Barrett said that, a year ago, that discussion would have centred on teaching people about prompting. But things have moved on.
“Now I kind of think that that’s just basic literacy. You have to know how to prompt AI as part of basic literacy.”
What is increasingly happening is that people who are learning new skills are seeing a real impact on their jobs: “Their careers are changing because they’re using AI,” she said.
WAN-IFRA Members can view Jane Barrett’s presentation on our Knowledge Hub.
When staff members were asked to define specific problems for AI to solve, and were then equipped with appropriate AI tools, the results were striking.
“We found that the skill most people develop very, very quickly is essentially development skills, software engineering skills, the ability to run code,” she said.
“People who’ve never written a line of code in their life – didn’t know Python was a language, not just a snake – are now producing really good code.”
“Is it fully scalable? It is not. But in terms of their own skills, they’re suddenly seeing huge new possibilities open.”
Structures for peer-to-peer learning
Barrett said that she had noticed a pattern when it comes to AI boosting people’s careers.
“Those people who have basic literacy, they’re growing, but they’re growing a little bit, in a linear way,” she said.
Meanwhile, “people who are really leaning into learning new skills, who are throwing out the fear, and who are saying, ‘I can do this myself,’ are growing exponentially. And that gap between people who are really leaning in and people who are just kind of pottering along as we’re asking them to, is really growing.”
Reuters supports skills learning by helping people learn from each other, Barrett said. The key here is to combine various, seemingly unrelated profiles.
“By taking people out of their silos, out of their roadmaps, and out of the typical ways of doing things, we’re pulling together individual teams, almost like squads, to work on something for a couple of months, giving them the tools and the time they need, and seeing what they can come up with,” she said.
“What we’re finding is that by opening people’s minds and networks to different people, people are learning much faster.”
As for internal communications about AI, Barrett said Reuters uses three main channels:
- A weekly newsletter for new AI tool suggestions
- Monthly town halls with updates on broader AI strategy
- Targeted groups (“AI champions”) in the newsroom, commercial and product engineering teams who meet once a month.
There are also specific training sessions on topics that are relevant across departments, such as how to use AI to analyse data.
The goal is to build communication opportunities “where it’s not just one-way communication, but it’s top-down, bottom-up, and then peer-to-peer,” Barrett said.
From knowledge transfer to mindset change
To provide some context for the current transformative moment, Barrett went back to the turn of the millennium, when “there was this thing exploding called ‘the internet.’”
Back then, many younger media professionals were excited by the internet’s potential – but “it was just crushed by a very old-fashioned leadership.”
That is why Barrett places strong emphasis on empowering people in her current role.
“A good idea can come from absolutely anywhere. So how can we create that environment where people feel valued, that everyone can have a great idea?” she asked.
Still, many people have concerns about AI tools, and high on that list for many is the prospect of their jobs being taken over by AI. An audience question asked how Reuters addresses those concerns among staff.
“You have to address it with both honesty and compassion,” Barrett said.
“The truth is, there are many of our colleagues who do more of the rote work,” she said – the kinds of tasks AI could potentially do just as effectively, or even more so.
“We’ve got to give people the honesty about the fact that AI is very powerful, that it might well be able to do some of the jobs that have been done by humans. And we have to have the compassion to help them through that fear, and help them move up the AI literacy scale and build new skills,” she said.
“Unfortunately, change is change, and some of us will end up with our jobs changing underneath us.”
Instead of fighting the change, the key is to “help people, empowering them to re-skill and upskill themselves, to learn how to use AI. So that whether it’s their job or it’s a different job, they are actually equipped for the future workplace we’re moving into.”
