There have already been several studies on how journalists use artificial intelligence (AI), but many of those studies interviewed relatively small samples of news workers so they can’t generalise about AI use more broadly. It’s true that a few larger surveys have now taken place. However, some mostly surveyed early adopters and others didn’t distinguish between current use and planned future use. So, we still have a limited understanding of whether and how journalists are using AI professionally. So, we decided to survey a representative sample of journalists — in the UK. We asked about their and their newsrooms’ actual use of AI and how they perceived and approached it. Our report, based on the survey, is published by the Reuters Institute. Here are some of our key findings:

How often is AI used? More than half (56%) of UK journalists use AI professionally at least once a week. Only 16% have never used it.

How is AI used? The most frequent uses monthly are for language-processing tasks: transcription (49%), translation (33%) and copy-editing (30%). Generating a first draft of articles (10%) and fact-checking (12%) with AI is less common. Only a small minority use AI for generating audio (4%) or video (2%).

Who uses AI? AI use is more prevalent amongst younger journalists, male journalists and journalists with more management responsibility. There are differences across beats (it’s more prevalent in business than lifestyle) but use is not associated with journalists’ contract type.

Has AI improved journalists’ job satisfaction? Those using AI more often are more likely to believe they work on low-level tasks too frequently and are not more satisfied with the amount of time they work on complex and creative tasks.

How do journalists perceive AI? Most UK journalists perceive AI as a large threat (62%) and only a small minority (15%) perceive it as a large opportunity. All groups are pessimistic, but less so those with more management authority and those who know more about AI or use it more regularly.

What do journalists fear about AI? The most prevalent ethical concerns about AI in journalism include the loss of public trust in journalism (with 60% saying they are “extremely concerned”), accuracy (57%), and the originality of journalistic content (54%).

How is AI used in newsrooms? A majority (60%) of UK journalists say that there has been some AI integration in their newsroom’s processes, although many of them describe integration as limited. However, most of them expect their outlet’s use of AI to increase in the future (63 percentage point difference between those that think it will increase vs decrease).

What AI guidelines and training do newsrooms provide? Over half (60%) of UK journalists say their main news outlet has established AI guidelines around at least one of the issues we asked about. The most common were ‘human oversight’ (44%), ‘data privacy and security’ (43%) and ‘transparency’ (42%). Only 32% say their outlet provides AI training.

Do newsrooms build their own AI tools? Only 9% of UK journalists say their main news outlet only uses tools developed in-house, with 57% saying their newsroom only uses third-party AI tools and 34% saying it uses a combination of both.

Does newsroom size matter? Journalists working for independent newsrooms report lower levels of AI integration, higher reliance on third-party AI tools, and expect less future AI integration than journalists working for conglomerates.

Many thanks to my fellow authors—Sina Thaesler-Kordonouri and Richard Fletcher—for their contributions to the report and also to the Reuters Institute team—especially Alex Reid, Eduardo Suarez, and Matthew Leake—for supporting the publication.