AI and the media: partners rather than rivals

The media has been in crisis since the rise of social media and platforms where content creators proliferate. As a result, readership and traffic have fallen dramatically. 

However, if there is one thing that the 2025 report by the University of Oxford and the Reuters Institute is certain of, it is that many people turn to well-known media outlets or government sources to verify information they find on social media. 40% of those surveyed by international market research and data analysis firm YouGov still trust news from general media outlets.

On the other hand, ‘chatbots and AI interfaces are emerging as sources of information, as search engines and other platforms integrate real-time news,’ although they remain low in numbers, according to the Oxford and Reuters report. 

Artificial Intelligence is based on searching for data on the internet, which is not always accurate or up to date. Nor can it access information that is accessible through personal accounts or paid subscriptions, according to programmes such as ChatGPT. 

However, some newspapers claim that these programmes also use their paid articles to generate responses. 

‘Automatic content generation tools, in many cases, do not accurately verify data,’ explained Ramón Salaverría, professor of journalism at the University of Navarra. 

The study by the Tow Center for Digital Journalism at Columbia Journalism Review stated that ChatGPT and Gemini, among others, get both sources and links wrong so convincingly that it baffles users. More than 60% of the responses from these programmes are incorrect. 

The same study corroborated that newspapers that claim their paid content is being stolen by AI are right, because they access restricted or paid content.

For Salaverría, this technology is a very useful tool for journalists as a complement. KPMG's report ‘Media: Trends Shaping a Future of Transformation’ corroborates this view. According to its data, 57% of those surveyed already use it and 37% plan to do so, while only 7% have no plans to do so. 

‘On the positive side, Gen AI allows news organisations to respond quickly to user demands, optimising processes and personalising the experience of readers, viewers and listeners,’ according to the KPMG report. 

In a sector that has been so commercialised since the end of the 20th century, where live and instantaneous content reigns supreme, AI would take care of the more routine processes as well as improving commercial practices, according to Andrés Rodríguez, president of the Association of Magazines and the Spainmedia group.

The KPMG report also indicated that this tool is very useful for personalising information for users. However, ‘there are voices warning of the risks associated with the coexistence of multiple narratives adapted to each user on the same event’. 

And given the landscape of social media supremacy and content creation indicated by the Oxford and Reuters study, according to the KPMG report, 70% of the media outlets surveyed would use this technology for this purpose. Sixty-three percent are more inclined to use it for data analysis and trend prediction. 

AI would also be used in newsrooms, and the KPMG report confirmed that it is based on a ‘collaborative creation model’ between people and machines. 

‘These technologies are expected to make news production cheaper (+29 net difference) and more up-to-date (+16),’ according to the Oxford and Reuters report. However, the document explained that the majority of the audience does not feel comfortable with its use in the media because it is believed to affect the transparency and reliability of information. 

For these reasons, the Madrid Press Association (APM), through its ‘2023 Report on the Journalism Profession,’ advocated for the need to regulate AI. 97% of journalists believed that its use would have ethical implications and would increase misinformation. 

Santiago Tejedor, researcher and director of the Communication and Education Office, concluded that Artificial Intelligence technology is extremely useful in routine tasks that can be automated, but it needs a human agent to review the content because it cannot distinguish between erroneous and truthful information. 

For a month, the Italian newspaper Il Foglio tested how effective AI is in a newspaper. So, they decided to create a print newspaper written entirely by artificial intelligence. 

The media outlet observed that technology can be ironic and even irreverent, but it can never replace flesh-and-blood professionals. Editor Claudio Cerasa found that AI itself highlighted what makes journalists unique by responding as if it were him writing: 

‘Simple: the idea. The obsession. The taste for deviation. The desire to understand where others are only summarising. The pathological inability to conform. The annoyance with banality. The enthusiasm for detail. The nose for phrases that seem neutral but mean everything.’