The rise of generative AI like ChatGPT and Perplexity has sparked my curiosity about the possibilities and limitations of this technology. Over the past year, I conducted three in-depth experiments, each in collaboration with experts from different fields: medical reporting with a physician, psychological applications with a psychologist, and text interpretation in journalism. In this blog series, I share key lessons from these studies. We begin with one of the most challenging domains: journalism. How can AI contribute to more reliable, transparent, and interactive news?
Why people are avoiding traditional media more often
In the Netherlands, distrust of traditional media (MSM) is growing. More people are turning away from established news sources and seeking information from alternative platforms. This stems from fundamental criticism of how media operate, select stories, and cover societal issues. The main points of criticism are:
Media Concentration: A Market Dominated by a Few
The Dutch news market is dominated by just two major media companies: DPG Media and Mediahuis. They control a large portion of newspapers, TV, and online media channels. Critics argue this concentration leads to uniform reporting and a lack of journalistic diversity. Alternative perspectives receive less attention, reinforcing the perception of a closed media landscape where dissenting voices are barely heard.
Political Bias: Journalism or Activism?
Consumers feel that reporting on issues such as climate change, immigration, and COVID-19 policies is often one-sided. Many news stories are seen as alarmist, offering little room for critical analyses or alternative views. Framing plays a key role: terms like “conspiracy theorists,” “climate deniers,” and “racists” are frequently used to discredit opposing opinions. This one-sided treatment strengthens the perception that the media follow social agendas rather than objective reporting, undermining trust and fueling concerns that MSM are engaging in activist journalism.
Sensationalism: Outrage Sells Better than Nuance
News organizations are often accused of sensationalism. Headlines are crafted to provoke outrage or fear rather than stimulate nuanced conversations. Social media has exacerbated this trend by forcing media outlets into a “clickbait race” for attention. Critics argue this comes at the expense of depth and context, creating an information climate where emotion takes precedence over content, hindering constructive societal debate.
Disinformation and the Monopoly of Fact-Checkers: Who Checks the Checkers?
An increasing number of Dutch people view the fact-checking process of traditional media as biased and selective. Media organizations often collaborate with fact-checkers who, according to critics, mainly target dissenting opinions rather than conducting broad research into complex issues. This selective focus creates the impression that fact-checkers primarily protect mainstream opinions, fueling mistrust of the media, which critics believe only reinforce widely accepted truths.
AI in journalism: improvement areas
Technological advances create opportunities that transcend the traditional boundaries of journalism. Generative AI like ChatGPT and Perplexity offers ways to address deep-rooted issues in the news industry. What seems speculative now could quickly become reality. How can AI contribute to a more reliable and balanced media landscape?
Media Concentration: A Market Dominated by a Few
The news market is controlled by a few large media companies, leading to limited diversity of perspectives. AI can expose this concentration by mapping ownership networks and systematically analyzing news articles for source similarities, framing, and content. In the future, AI could monitor media platforms automatically for diversity and balance, uncovering hidden structures and revealing new niches otherwise invisible.
Political Bias: Journalism or Activism?
The perception of one-sided news reporting is a major cause of growing distrust in the media. This issue stems not only from obvious preferences but also from subtle framing strategies and selective coverage. AI can play a key role by going beyond superficial sentiment analyses. Technologies like Natural Language Processing (NLP) can analyze tone, context, and terminology used in news articles to uncover structural patterns of bias.
Platforms like Ground News compare news articles from different media outlets to identify political bias, showing how the same story is presented across left-wing, right-wing, and neutral sources. AI could extend this concept by performing deeper analyses, detecting framing through charged terms like “conspiracy theorists,” “climate deniers,” or “agitators.” This technology could automatically flag when specific words or expressions are systematically used to strengthen or undermine certain viewpoints.
Reducing Sensationalism: News or Spectacle?
The fight for attention has led to the rise of sensationalist news. AI can detect clickbait by recognizing sensational headlines and emotionally charged content. In the future, AI could rank news based on societal relevance rather than popularity, giving more visibility to quality stories and enabling consumers to make better-informed choices.
Disinformation and the Monopoly of Fact-Checkers: Who Checks the Checkers?
With the explosion of disinformation, fact-checking has become essential. AI can already automatically verify sources and scan content for factual inaccuracies. However, who oversees the fact-checkers themselves? AI offers a chance to make the process more transparent by generating real-time reports on which sources were checked and how often corrections were applied. This could democratize control over disinformation, extending it beyond a handful of institutions.
Conclusion:
The combination of media concentration, political bias, sensationalist reporting, and the monopoly of fact-checkers fuels growing distrust in Dutch media. AI has the potential to unravel these issues through advanced text analysis, semantic recognition, and sentiment analysis when applied transparently and ethically. Whether one supports AI or not, experience shows that it will make its impact where it matters most.