Year after year we look at studies and forecasts of new trends in the field of communication, not so much to obey what they say, but to take ideas in the face of challenges and to broaden the vision. If 2023 was the year of the social presentation of Artificial Intelligence1, we received in this same section that last year would be the year of reaction to AI. To begin with, we will analyze the degree to which these expectations have been met and, with it, we will bring here some readings and analyses that come from the international community by 2025.
What the 2024 left in this area and what this year will bring is appropriately compiled in the Reuters Institute’s Journalism and Technology Trends and Predictions 2025 study. The report, based on surveys conducted by authors Nic Newman and Federica Cherubini to 326 media managers (most of whom are from Europe but there are from all continents), explains that the media is intensely concerned with the profound impact of creative or generative artificial intelligence (linked to the footnote above, this term seems to be expanding on a social/commercial scale) on journalism and the media business. The principles and guidelines associated with AI are already largely in force in some major newsrooms and have also created roles for AI specialists.
These activities are not limited to newsrooms, as this is what members of other areas, such as the commercial area or the content and products, are looking for. In fact, improving efficiency in automations for internal tasks is the task that attracts the most interest and attention, according to those responsible, and this will be the main work related to AI for this year, along with: refining the capacity for personalization and recommendations; deepening the creation of human-supervised content; and developing news gathering functions (for example, for fact-checking, data journalism or investigations). It is therefore a matter of taking further steps along the path that has already begun, as the toolboxes developed on their own are emerging here and there, with applications and programs already in operation. The JP/Políticas Media Group, one of Denmark"s leading groups, for example, has created the MAGMA assistant for writing and editing, which produces full-text drafts with simple words and data and is used by three different publishers of different styles. The Finnish Helsinging Sanomat, for example, has its own tool to help you with research and reporting, capable of automatically sorting data and events and creating timelines, among other things.
The use of AI for the Spanish media Datadista, specialised in data journalism and research, reduces the work of weeks or months to a few days thanks to new scattered sources and models of data processing. Fact-checking processes are also accelerating dramatically with the help of AI, especially in the case of videos and audios, as confirmed by the Quint media in India; the German weekly Der Spiegel has also developed its own tool aimed at automated and routine verification, with the intention of adding more features later on.
The problem of the depreciation of these high economic costs on the other side of the coin of this technological development is also among the concerns of these managers. It is not clear whether this improvement in the ways of working and in the content will lead to an improvement in the business model. There is little doubt, however, in asserting that the emergence of the creative AI is, at least in part, transforming newsrooms. And it is clear that the transformation will continue if we look at the next major moves towards audiences. Media officials say they will deepen their efforts to convert texts into audio; the automatic summaries at the top of the articles will be expanded; they will integrate chatbots/AI search functionalities to interact with the audience (Aftonbladet, the Financial Times or the Washington Post have already begun to break this path); and some will run tests with tools to convert texts into video. But, as the authors of the report predict, not everything will be done as desired by those responsible for the media, since the audience itself, ordinary users, will be able to adapt their content in the same search engine with new search engines such as Dia2 (based on AI, this new model that seems to take search engines to another level will appear in 2025).
This puts another question on the table: to what extent should the media develop their own tools in view of the fact that other technology developers also advance with applications that perform similar functions. That’s how they’re integrating free extras like Google, Microsoft, Zoom, or Dropbox for text summaries, transcriptions, translation, or data organization. Tools such as OpusClip AI will emerge to mold not only texts, but also videos, taking a long conversation and automatically presenting them as if they were made specifically in this short format and full of effects language for Tik-Tok.
With all these tools available to us, the amount of synthetic content that has emerged this year has raised some alarms and, in this sense, it is interesting that the authors highlight the concern that could lead to the collapse of the model3 itself. In other words, if these synthetically produced low-quality contents begin to be a model for AI (LLM, etc.) and become a fodder for its training (because they are present in the network itself), they will enter a spiral, increasingly distant from original sources, on their way to collapse. Essential to avoid this will be the protocols for setting metadata criteria (such as C2PA), which will function as a certificate of the origin of the contents and help the algorithms to decide what to prioritize.
Another challenge that this transformation poses to solid and stabilized media is the retention of the working group. On the one hand, the extended model of individual productions and influencers, thanks to the possibilities offered by the tools, and on the other hand, the detailed analysis of the audience allowed by digital analytics, has led to the emergence of individual communicators, and in some cases, those considered as stars have escaped, as happens in football, through signings. The report by Newman and Cherubini shows that there are different opinions among those responsible for the media, so that while some believe that they should build their own influencers in the media, others believe that being linked to the brand, the team, still has value. However, there is a greater challenge if we look at the more technological area of the team: most managers agree that in this transformation of newsrooms it is more difficult for them to attract or retain data analysts and software engineers. Looking ahead, it is expected that the negotiations will be more complex in terms of retaining specific members of the Working Group and that the media will be involved in the ongoing development of the relay. It is also a great opportunity for communicators interested in technology, because as the hybridization of newsrooms progresses, the combination of these two skills will prevail.
The pressure on the stabilized or traditional media is not small for 2025, if we add to the development and transformation of the technology mentioned so far the repercussion that is having a reactionary global political environment (insisting on questioning the credibility of the traditional media). According to the conclusions of the Reuters Institute’s report, audiences, especially young people, are attracted to platforms such as YouTube and TikTok, making it more difficult for traditional media to connect directly with this audience, which in turn fuels the model of individual or personalist communicators and the competition between them. On the other hand, media managers will pay special attention to making media and products more attractive and deepening the personalization of the audience, taking advantage of AI but also focusing on the human nature of the content that AI will never achieve, whether it be in reports, creative works or experiences of the audience. Regarding the economic aspect, they foresee the reduction of costs, the diversification of income, and the greater integration of information with contents related to lifestyle; this latest trend, with the intention of strengthening loyalty and habit. In this connection, one of the main tasks of the year will be to redefine the role and value of the media in the era of polarization, misinformation and overdose of content, taking into account both the working groups and the audience itself.
The references
In Jordan, M. I. I. I... (2019). Artificial Intelligence—The Revolution Hasn’t Happened Yet. Harvard Data Science Review, 1(1). https://doi.org/10.1162/99608f92.f06c6e61
According to Newman, N., In the case of Cherubini, F. (2025), Journalism, Media and Technology Trends and Predictions 2025. Reuters Institute for the Study of Journalism. https://doi.org/10.60625/risj-vte1-x706
More places to stay in Newman, N. (2024), Journalism, Media and Technology Trends and Predictions 2024. Reuters Institute for the Study of Journalism. https://doi.org/10.60625/risj-0s9w-z770
1 Regarding the term and concept of Artificial Intelligence, there are critical voices (Michael I. Jordan, for example) who say that what we call AI is very far from true Artificial Intelligence, generally speaking of Machine Learning (ML) or Machine Learning, or activities corresponding to disciplines such as Intelligence Augmentation (IA) and Intelligence Infraestructure (II); a new engineering system, in short, but not artificial intelligence. There is also an academic and commercial conflict.
https://hdsr.mitpress.mit.edu/pub/wot7mkc1/release/10
2 https://www.diabrowser.com/
3 https://www.nature.com/articles/s41586-024-07566-y