This year’s research by the Reuters Institute for the Study of Journalism shows that as newsrooms integrate AI tools into journalists’ daily routines, audiences feel fear and distrust.

More than half of the respondents in the United States and over 60% in the United Kingdom say they would be uncomfortable reading news that is largely produced by AI. Audiences are most receptive to the use of AI in media areas where it can improve their experience of accessing and consuming news, such as through personalized news feeds. A significant portion of the audience opposes the use of AI in creating content on sensitive topics and generating realistic-looking images and photos. Only 36% of respondents in the US, UK, and Mexico feel comfortable consuming human-generated news with the help of AI, and an even smaller proportion (19%) feel comfortable consuming news that is mostly AI-generated but with human oversight. Most respondents believe that AI should be supervised by humans and that full automation should be prohibited.