#drdavidryfe

Regional journalism experts appear optimistic generative AI can help their snowed-under newsrooms.

By João Santos for Press Gazette 

A WAN-IFRA survey has found that half of newsrooms are already deploying generative AI in some form or another.

Large language models quickly garnered the attention of Reach, Buzzfeed, Conde Nast and other publishers across the globe over the past six months for its potential to streamline editorial processes.

In particular the free-to-access ChatGPT has fast shown its potential to reinvent the way content is produced and journalism is done. But in local news outlets hard-pressed for time and staff, is the adoption of AI likely to be an opportunity or a threat?

AI vs the humans in local journalism

A number of UK regional news publishers have already realised that AI-generated news stories could free up journalists to concentrate on more meaningful reporting.

Duncan Williams, the managing director and proprietor of Pulman’s Weekly News in south east Devon, said: “AI has got huge advantages to help journalists, particularly in the regional sector, but it’s not yet at a level where it’s going to be able to write nitty-gritty stories.”

Rather, he said he believes the more in-depth stories produced by human journalists will carry increased significance. He wrote an op-ed in his newspaper last month emphasising the importance of human bylines for trust.

“As an editor, I would not dream of publishing a story by someone I did know the name of or who was not a trusted and credible human source,” he said. “Bylines not only protect my readers, but they also hold journalists accountable for their work.”

Dr Joy Jenkins, an assistant professor of Journalism at the University of Tennessee in Knoxville, has worked extensively on how new technologies can be sustainably introduced into local industries.

She said: “AI can be used on more basic types of reporting in order to free up journalists to do the more investigative, in-depth types of stories they’d like to do but don’t always have the time and resources to produce.”

Eliz Mizon, a media writer and communications lead at Bristol Cable, believes AI could help local journalists carry out important “institutional accountability and community cohesion” roles, potentially reversing a trend that has seen around 300 local newspaper title closures since 2005 and left others hanging by a thread.

Will AI help or hurt regional journalist numbers?

A recurring worry that crops up in discussions of AI and news is whether the technology will end up replacing journalists, particularly in light of the economic headwinds that have encouraged many publishers to lay off staff in the past year.

Jody Doherty-Cove, Newsquest‘s special projects editor, is currently at the forefront of developing, coding and implementing AI tools for over 180 newsrooms across the UK, having just helped the Gannett-owned publisher establish its first AI reporter positions.

He explained: “A common misconception in the public discourse is the notion that AI, like some malevolent force, is poised to supplant human journalists in a ruthless bid to cut costs.

“Local journalism is a craft that requires investigation, relationship building, and nuanced understanding – elements that no machine, no matter how sophisticated, can truly replicate.”

Dr Jenkins agreed, explaining that AI should be seen as a “really affordable and helpful” tool which, if used properly, can be an “extremely effective way for local journalism to maximise its resources”.

Pulman’s owner Williams added: “It would streamline both the numbers of staff required to run our titles in an economic fashion and also the actual revenue streams that are the lifeblood of local journalism.”

This is the approach that Kallum Gethins, managing director at Dorset News and editor at View From Weymouth, has taken when incorporating AI into the running of both publications.

“If we find a story, we gather the facts, make a short sentence and then bring that into the AI software which produces a three to four-paragraph article, we proofread it and publish it,” he said. “The whole process takes about ten minutes.”

Much like the advent of social media and citizen journalism, AI could therefore lower the barrier of entry into the industry, making quality and consistency more affordable.

Gethins said: “I think that all organisations should implement AI because it could help their business massively, especially if the organisation is brand new and can’t afford journalists. For instance, we probably won’t be using AI-originated content in the long term and will look for potential journalists to join our franchise.”

The challenges ahead for AI in local journalism

As automation seeps into the media industry, the utility of AI in local journalism could be as far-reaching as the industry demands.

As well as editorial tasks, Dorset News’ Gethins has begun using different AI software engines for administrative work, for example producing email templates suited to potential investors and creating budget plans.

Automation could magnify the results of advertising campaigns and help integrate local papers more fully into social media, something which, Dr Jenkins explained, has been a historical struggle for many publications.

However there are considerable challenges to overcome before AI can be fully integrated into journalism in a way that is both ethical and practical.

Doherty-Cove said: “We must ensure that AI-generated content does not escape the watchful eyes of human editors who, with human oversight and rigorous fact-checking, can ensure AI-generated content meets their exacting requirements.”

Chatbots such as ChatGPT rely on what the user feeds into it, as the data on which they were trained is in some cases out-of-date. This means the fundamental newsgathering framework which leads to a fully-fledged story is, for the moment, outside the purview of generative AI programmes.