12/17/2025 | News release | Distributed by Public on 12/17/2025 01:36
2025-12-17. If we want to defend reality itself, we must embrace AI while doubling down on original reporting, writes Peter Vandermeersch.
by WAN-IFRA External Contributor [email protected] | December 17, 2025
By Peter Vandermeersch
Journalism did not suddenly reinvent itself in 2025. But it was the year in which long-looming forces finally made themselves impossible to ignore. Generative AI surged ahead.
Tech companies and publishers clashed openly. Trust in news continued to slide. Young people drifted further from traditional news formats. And political actors stepped up their attacks on the information order. Together, these pressures pushed journalism to a decisive crossroads.
Journalists still tell stories, but the craft is shifting fast. The job is no longer simply to publish information - there is too much of it, arriving too quickly, from everywhere. Instead, journalism increasingly means interpreting, filtering and verifying a torrent that threatens to overwhelm public understanding. Information overload is no longer a metaphor; it is the daily condition of modern life.
Against this backdrop, the struggle between publishers and the major AI developers erupted into view. It became clear that the language models shaping our digital world - GPT, Claude, Llama, Gemini - were trained on vast swathes of journalism, usually without permission.
When tech executives told the US Congress their models rely on "publicly accessible web content," publishers heard an admission of something long hidden: our journalism has become raw material for someone else's business model.
As Axel Springer CEO Mathias Döpfner put it: "Quality journalism is the foundation on which AI companies build their products."Undermine the foundation, and sooner or later the entire structure collapses.
The media industry has been divided on how to respond.
Some outlets have signed licensing deals; others, led by The New York Times, have taken AI companies to court, arguing that industrial-scale scraping amounts to industrial-scale copyright theft.
European publishers have joined out of fear that reporting will be hollowed out while AI companies reap the value created by others.
The core issue is simple: if news organisations want to keep funding reporting, verification and investigative work, then AI must help pay for the journalism it depends on.
The enormous value generated by AI systems does not arise from nothing; it comes from decades of journalistic labour. If AI companies want ongoing access to that knowledge, they must help sustain the ecosystem that produces it.
But the real transformation is happening inside newsrooms. And here lies the paradox of our moment: AI is both the technology that threatens journalism and the technology that could strengthen it.
Used well, AI can take over the tasks that suffocate reporters - sorting documents, summarising dense reports, transcribing interviews, comparing sources, spotting patterns.
These jobs are essential but time-consuming, repetitive and rarely what makes journalism distinctive. If AI frees journalists from these burdens, it gives them more time for the work only humans can do: developing sources, making judgements, reporting on the ground, challenging power.
Yet fear persists. Some journalists worry AI will corrode quality; others that audiences will no longer know what's human and what's synthetic. And many fear their employers will use AI to cut jobs. These anxieties are understandable. But they miss the greater danger.
The real threat is not AI inside newsrooms - but AI outside them, spewing out unlimited synthetic content that looks like news but lacks any journalistic scrutiny.
This is why transparency is now essential. In an era of deepfakes, fabricated quotes and synthetic narratives, journalism must show not only what it reports but how it knows. Transparency is not a branding exercise; it is the only route to rebuilding authority.
One form of journalism remains irreplaceable: Investigative reporting. It generates the one thing no machine can produce - new facts.
AI can analyse and summarise, but it cannot expose corruption, obtain a confidential document, build trust with whistleblowers or confront a minister with uncomfortable truth. That work remains human. And the best investigative reporters will use AI not as a threat but as a power tool.
Meanwhile, audiences are changing as fast as the technology.
Younger people are turning away from legacy news in favour of individuals who explain events in short, direct, visual formats.
France's Hugo Travers (HugoDécrypte), Spain's Emilio Doménech and the phenomenon known as News Daddy now reach millions.
For many young people, these creators are not an alternative to journalism but a replacement - less formal, more immediate, more human.
And then there is the political climate. In the United States, 2025 saw the most intense pressure on the media in decades. The White House launched an official site inviting citizens to report "fake news" - essentially a modern blacklist.
Donald Trump sued The Wall Street Journal for at least $10bn for over reporting on a supposed letter to Jeffrey Epstein and threatened similar actions against The New York Times and the BBC. These cases may have little legal merit, but they serve an unmistakable purpose: to terrify the press into silence.
A president who seeks not just to vilify the media but to bankrupt it strikes at a democratic cornerstone.
This is the world journalism now inhabits: Threatened by tech giants who extract its value, by AI systems that can replicate its surface without its substance, by audiences who distrust or ignore it, and by political actors who brand it an enemy.
In such a moment, the answer is not less journalism - but more. Not defensive journalism - but better.
We need journalism that understands technology rather than fears it, that embraces transparency as a source of credibility, that collaborates with creators who reach new generations, and - above all - that keeps doing what no machine can do: uncovering what is true.
About the author: Peter Vandermeersch is a former editor-in-chief of De Standaard and NRC Handelsblad; from 2019-2025 he was the CEO of Mediahuis Ireland, publisher of the Irish Independent and The Belfast Telegraph.
He is now Mediahuis Fellow: Journalism and Society.
WAN-IFRA External Contributor