“Many journalists in the Global South are already using AI and they want guidance on how to do this ethically”

There is no hotter industry topic of conversation than the use of AI by journalists and within newsrooms. Yet much of that discourse is informed by AI use in Western news outlets. Just as AI tools trained on Global North sources or in English may display bias, a lack of understanding how journalists in the Global South and emerging economies are using AI is skewing the discussion.

A new report from the Thomson Reuters Foundation wants to change that. Based on 221 responses from journalists in more than 76 countries and territories in the Global South to a survey conducted in the last quarter of 2024, the report explores the barriers to AI use, where AI is already changing how journalists work and where journalists are optimistic about AI’s impact.

I spoke with the report’s author, Damian Radcliffe, to find out more about the uses and perceptions of AI in journalism in the Global South’s newsrooms. Our conversation has been edited for length and clarity.

Q. Why does this report focus on AI use by journalists in the Global South?

A. Generative AI is evolving incredibly quickly and it’s fair to say that a lot of existing studies have focused on the Global North. Even within good global studies, the sample sizes inevitably tend to skew more Western. That’s something we wanted to address through this study. We wanted to elevate the voices of journalists in the Global South and be able to hear directly how they are using these tools and their fears and concerns.

It felt like there was an opportunity to paint a picture of how AI is being used by journalists and newsrooms in the Global South, but also to ask what that means. There are clear recommendations at the end of the report – for journalists, newsrooms, funders, policymakers, and platforms – to translate this research into actions that could make AI use in the Global South more equitable and help upskill journalists in those regions.

Q. What stood out to you from the responses of the journalists?

A. There is a disconnect between the number of people who are using this technology and the amount of training, support and active encouragement from newsroom leadership. We found that there’s not a lot of training going on. Most journalists are self-taught [when it comes to using AI tools] and have learned by playing with tools, participating in online programmes or watching YouTube tutorials.

There’s not a lot of formal training and very little guidance in terms of AI policies within newsrooms. [Only 13% of those surveyed said their newsroom had an AI policy.] This means there’s a lack of clarity for people within newsrooms and individuals could be pursuing their own way of doing things, which may or may not be ethically or editorially sound. It’s also potentially confusing for audiences, because they have no indication of if AI is being used.

There were high levels of AI use, but also high levels of concern. For example, about how AI tools can be used for misinformation or how AI use by journalists could impact the quality of journalistic output or make journalists less creative. There’s enthusiasm for uses such as AI for transcription, translation, summarisation, brainstorming and so forth. But there’s less adoption of more sophisticated use of AI [such as AI for data analysis or audience engagement products]. 

Cost of access [to tools] is a legitimate concern in many newsrooms and there is often a lack of training, awareness or understanding of some of the more sophisticated ways of using these technologies, such as using it to generate or edit video. We also have to recognise that some of those more sophisticated uses are just not part of what people do day to day. I suspect this will change over time as these tools get better, if we're able to improve access to them and as newsrooms continue to shrink.

Q. Did any of the attitudes expressed towards AI surprise you?

A. A lot of respondents seemed quite optimistic about regulation. There’s recognition that regulation is not a panacea for AI’s risks. But given the range of concerns expressed, there seemed to be a lot of hope that regulation could address many of those, from issues of access, bias in data sets, inconsistent use guidelines and concerns around transparency regarding AI use within journalism.

We also asked about journalists’ awareness of AI policies and regulation in their country and worldwide, and the levels were low. Even if you make some regulatory headway on these issues, how are these journalists going to find out? How are these newsrooms going to incorporate them?

Q. Did your research give you a sense of the impact of AI on journalists working in restricted regimes or where journalism is under threat?

A. There are examples of virtual anchors “presenting” the news in India and Venezuela, something that’s been the case for some years in China. There’s a media freedom perspective to this, which is around using these AI tools as a way to protect the identity of journalists whilst sharing the news. I wonder if we will see this in countries where there are increasing encroachments on media freedom. 

The use of AI tools, particularly chatbots and avatars could be an engaging way to get the news out there, while providing journalists with a cloak of anonymity. That idea was new to me.

Respondents to the survey raised clear concerns around AI’s impact on trust in journalism. That’s a universal concern. But in countries where journalism is already under threat, AI becomes another potential means through which trust can be undermined. This could be through AI creating inaccurate information that journalists in a hurry might not check as well as they should; or because AI enables false information to be easily created and distributed by anybody, which was a concern more prominent amongst these journalists.

Q. What should newsroom leaders in the Global South take from this report?

A. If you are a newsroom leader, the biggest takeaway is that so many of your journalists are already using this technology and they want guidance on how to do this in a way that is ethical and in line with best practice – but they don’t know how to find that or what that looks like. We need more case studies from the Global South, so that journalists can see how newsrooms like theirs are using this technology, newsrooms that don’t necessarily have deep pockets or access to other forms of support, such as their own legal teams.

Q. Your report makes a few recommendations for how AI can be used responsibly. What role can journalism funders and media development programmes play in this?

A. Many journalists are already using this technology, but they're doing so unaided. Support through more formal training would be very welcome, in a way that can create a sense of community and best practice amongst practitioners. This would be a great outcome and a relatively easy one for funders to support.

Freelancers are an important part of this conversation. Where formal AI-related training does exist, it's often expensive or tied to particular newsrooms. Even if you’re a freelancer for that newsroom, you probably won't get invited to the training. The cost of not working for a day to attend training can also be prohibitive for freelance journalists. I’d love to see what more can be done to support freelancers in the Global South to ensure they are not left behind.

Q. What should journalists and newsrooms in the Global North learn from this research?

A. There can be misplaced arrogance in parts of the Global North that this is where the most impactful and innovative journalism is taking place, but there’s lots to learn from other regions. 

Journalists in the Global South are tackling many of the same fundamental challenges: sustaining business models, using technology in smart and ethical ways, rebuilding trust, combating news avoidance and finding fresh ways to engage younger audiences. They’re often doing it with far fewer resources and in more challenging, sometimes hostile, environments.

I would encourage more newsroom leaders in the Global North to broaden their gaze, recognise this creativity and understand the opportunities to learn from their peers worldwide. Journalists will find inspiration and innovation everywhere, if they know where to look. 

The report, the first in the new TRF Insights series, is titled 'Journalism in the AI era: Opportunities and challenges in the Global South and emerging economies' and can be downloaded in this link.

Editor's note: The author of this piece has worked as a consultant for the Thomson Reuters Foundation, but has not been involved in the writing of the report covered in this piece.