A recent survey of roughly 12,000 adults across Argentina, Denmark, France, Japan, the United States, and the United Kingdom reveals that people continue to place greater trust in news produced primarily or entirely by humans than in content generated by AI. According to data published by the Reuters Institute for the Study of Journalism, trust increases in proportion to the level of human oversight.
While the public generally feels comfortable with how they believe GenAI is currently used in journalism, concerns persist around its application to certain news-related tasks. Optimism about GenAI’s future varies by sector, with news and politics standing out as areas of skepticism.
The study also shows a growing adoption of GenAI tools, with trust in these technologies rising alongside user familiarity. As this trend continues, the advantage currently held by human news professionals may diminish, especially if news leaders fail to actively reinforce public confidence in the value of human-driven journalism.
GenAI in news and journalism
The percentage of those surveyed who are more comfortable with news produced entirely or mostly by human journalists has risen slightly compared to data from the previous year, while trust in news generated primarily or entirely by GenAI has fallen slightly. This indicates a growing preference for human leadership in news, presenting an opening for news executives to bolster public confidence. Across all six countries studied, a strong preference for human oversight in news prevails.
- 62% are comfortable with news made entirely by humans.
- 43% are comfortable with news produced mostly by humans with help from AI.
- 21% are comfortable with news produced mostly by AI with some human oversight.
- Only 12% are comfortable with content generated entirely by AI.
Comfort levels vary according to how AI is being used. Most participants are fine with GenAI use in checking grammar, spelling, and providing translation. They are less approving of use for research, writing, and data analysis. The public is decidedly disapproving of GenAI tools being used to rewrite content for different audiences, generate a realistic image when a photograph isn’t available, or create an artificial presenter or author.
Fortunately, people’s comfort level with journalists using GenAI for certain tasks is aligned with how often they think journalists are already doing so. It appears that most of those studied believe journalists are using AI in ways that they find acceptable, and few believe it is commonly used in the ways they would find most unacceptable.
While there are differences among the countries surveyed,news is trusted significantly more than the most widely used and trusted GenAI system, ChatGPT, in almost every country studied.
- In Denmark, 72% report trusting news while only 32% report trusting Chat GPT.
- In Japan, 60% trust news; only 31% trust ChatGPT.
- In the UK, 45% trust news; 20% trust ChatGPT.
- The USA and France have lower margins, with 36% of respondents from both countries reporting trust in news while 27% trust ChatGPT.
- Only Argentina reversed the trend – with 37% trusting ChatGPT- more than the 31% who reported trusting news.
It’s worth noting that ChatGPT was found to be the most trusted GenAI tool among survey participants. This means that the differences in trust levels would be even more stark if comparing news with lesser known GenAI tools.
Caution: Trust in GenAI grows with familiarity
Increase in regular use of GenAI tools is leaping rapidly. The proportion of survey participants who reported having used a standalone GenAI system such as ChatGPT rose from 40% in 2024 to 61% in 2025. Those reporting weekly usage nearly doubled in a year, jumping from 18% to 34%. So, if trust rises with use and familiarity, traditional news media could soon lose their edge in public trust when compared to GenAI tools.
Not surprisingly, younger generations were found more likely to both use and trust GenAI tools. 59% of people in the 18–24 age range reported having used any GenAI tool in the last week, compared to 20% of those aged 55 and up. However, that age gap is driven mostly by ChatGPT. Other GenAI tools, including Google’s Gemini, Microsoft’s Copilot, Meta AI, and Grok, had narrower use differences across age groups. This is probably because the later tools are embedded within widely used products.
Optimism about GenAI varies by use case
The public generally leans optimistic about the future of GenAI. Across all six countries studied, on average 29% are optimistic and 22% pessimistic about the impact of such technologies. However, the share of optimism versus pessimism varies by several factors. Pessimism outweighs optimism when it comes to GenAI use in news media, government, and politics. Optimism outweighs pessimism when it comes to GenAI use in health care, science, retail, and search engine efficiency.
- Only 18% think GenAI will improve their experience with political parties or politicians.
- 27% believe GenAI will enhance their interactions with news media.
- 37% believe GenAI will improve their interactions with health care professionals and scientists.
- 43% believe that GenAI will enhance their experience with search engines.
The lower levels of confidence in GenAI’s impact on politics and news suggests a perceived link between those areas. It also aligns with the stated preference for human-produced news.
The future of Gen AI and news
While this data provides some reassurance for news media leaders that the public values human news professionals, it also points to some areas of concern. As use of GenAI rises, so do comfort levels and trust in the utility of the tools, potentially eroding the advantage currently held by human journalists and producers.
To stay ahead and remain better trusted by audiences, news organizations should prioritize original reporting. They must differentiate their offerings from GenAI content and actively communicate these unique values to their audiences.











