Reaching and monetizing readers are the most important uses of AI systems in news, says a global survey of journalistic leaders. The REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM annual survey reports responses on several issues facing news organizations in 2020. The section on AI in journalism is one of six in the study, occupying about two-and-a-half of its 39 pages.
Under the heading ‘Doubts Emerging over the Role of AI in the Newsroom,’ the study says ‘automated recommendations’ (of stories) will be the most important use of AI systems in the year ahead, followed by ‘commercial uses’ and newsroom support functions, such as grammar-checking, tagging, and transcriptions. Applying AI tools for news gathering and automated writing will be the least important.
The results indicate AI is being seen as a business tool more than as a way to extend or create new methods of journalism.
- There is a general sense of optimism that machine efficiencies are creating more time for human journalistic work.
- Concerns are related to a disparity between ‘have’ and ‘have-not’ newsrooms, an ‘over-estimation of the capabilities of AI today,’ and the need for continued vigilance about the potential of AI systems creating and spreading fake news.
- RISJ canvassed 233 people from 32 countries. Of these, 93% responded to the questions about AI.
- The results are presented as a ‘strategic sample’ rather than being representative of the whole industry.
The section ends with bass notes about the gap between of AI’s place in society and the capabilities to cover it as a story.
‘The application of AI and the implications for privacy and democracy will be a defining issue not just of 2020 but of the next decade. But many worry that the technology will run ahead of our ability to shine a light on the way these systems are being programmed or used.’
Journalism, Media and Technology Trends and Predictions 2020
AUTHOR
Nic Newman – Senior Research Associate, RISJ
SEE RELATED
- Six trends you shouldn’t miss for 2020: from subscriptions to podcasts and AI – RISJ blog post
- Download report PDF – AI sections begins on page 22