The fight over public information
Technology companies are reshaping the information space in their own interests. Their AI tools can spread falsehoods, while systems that summarize publications may draw audiences and revenue away from the outlets that produced the original work. The stakes are not just commercial: journalism helps establish facts and sustain a democratic public sphere, and the question is whether the public can defend that role.
The harms of the digital economy extend beyond journalism. It collects personal data, manipulates public opinion, destroys jobs, restricts civil liberties and damages the environment. The article argues that its interests also align with those of the world’s most authoritarian and reactionary states.
Journalists of the Futo generation have lived through four major digital revolutions: the internet, social media, smartphones and now AI. The latest changes the relationship with truth. For technology companies, the boundary between truth and falsehood has disappeared.
AI answers are not reporting. They are words assembled through logic, the article argues, and too often they contain errors, distort facts or misinterpret information. Journalism, by contrast, depends on gathering evidence, checking claims and deciding what matters.
What reporting does that AI cannot
That work involves persuading sources to talk, collecting eyewitness accounts, comparing evidence and establishing facts. It also means adding context and nuance, and pursuing information that people in economic and political power may prefer to hide or ignore. Generative AI, the article argues, reveals nothing: it repeats and compresses material that has already been published.
Reporting can carry costs that are difficult to measure. Journalists covering war risk their lives. A recent Mediapart investigation into Jordan Bardella, leader of Marine Le Pen’s far-right party, brought anonymous cyberbullying and death threats against the newsroom.
Media organizations, meanwhile, answer for what they publish to readers and courts. AI companies appear to answer primarily to shareholders, and refuse to guarantee the reliability of their conversational systems in order to avoid legal liability.
If publishers’ work continues to be appropriated without permission or compensation, the article warns, journalism itself could disappear. The loss would reach beyond publishers: it would mean losing a public sphere grounded in free, pluralistic and democratic debate.
Where the public has leverage
The companies behind AI also depend on the information they are depleting. Without high-quality source material, they risk producing only low-quality AI-generated content. OpenAI has acknowledged that it is impossible to train modern advanced AI models without copyrighted material. A Microsoft vice president has also acknowledged that training on high-quality material significantly improves model answers.
Some publishers are already turning to the courts, including The New York Times. Its publisher, A. G. Sulzberger, recently said the litigation had cost the newspaper $20 million. Courts are also considering cases in France, including one brought at the initiative of an organization responsible for collecting payments for press-related rights adjacent to copyright.
The article’s case is compelling, but I think the unresolved issue is how that leverage becomes durable protection. Lawsuits can challenge appropriation, and public pressure can demand new rules; neither alone guarantees that independent reporting will remain accessible or financially viable.
Journalists also have to show the public what their work makes possible: holding power accountable, driving change, explaining a rapidly changing world and supporting civic participation. That is not simply a defense of publishers. It is a defense of who gets to establish the facts on which public debate depends.
Demonstrators protest outside an Anthropic office in New York, 14 October 2026. Photograph: Sarah Yenesel/EPA
Source: theguardian.com
Daily AI news
Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.
Only what matters — every day
Follow on X