It’s strange, isn’t it, how the most fascinating dinner-party topic of our lifetime turned out not to be football, or politics, or even the weather, but artificial intelligence. Nobody saw that coming. A few years ago, if you mentioned algorithms or neural networks, eyes glazed over; today, mention ChatGPT or Gemini and suddenly everyone has a story. Maybe it’s a trick, like asking the chatbot to write a poem about a traffic jam, or a genuinely moving piece of code that helped a friend draft a difficult letter. Or maybe the conversation turns to bigger things: the global race for AI dominance, the fear of mass unemployment, the enormous energy appetite of data centres, the possibility of AI-driven medical breakthroughs, and the terrifying transformation of warfare. These are huge, world-shaping subjects, and they deserve our attention. But here’s something curious, and frankly unsettling. For all the excited talk about what AI can do, we rarely talk about what it has already started doing to the very idea of truth. We seem oddly silent about the fact that this miraculous technology is also the most powerful engine of misinformation ever built. That silence, more than anything the technology itself does, should worry us. Because ignoring misinformation is itself a symptom of the disease: a quiet, collective shrug that says the truth no longer matters as much as it used to, and perhaps never will again. And that, in its own way, is exactly what misinformation wants us to believe.
Let’s be clear about the scale of the danger. The World Economic Forum’s Global Risks Report now places AI-driven misinformation among the most severe short-term threats to global stability. That’s not an exaggeration from a tech-sceptic think tank; it’s a stark warning from one of the most influential global institutions. To understand why, you only have to look at what’s become possible in the last few years. Artificial intelligence can produce fake images, videos, audio, and articles at a speed, cost, and volume that were unimaginable a decade ago. What once took a skilled graphic designer, a sound engineer, or a video editor days to create can now be done by anyone in minutes, for free, with a smartphone. And it’s not just about creating a single fake image; it’s about personalization. Algorithms can tailor misinformation to the psychological profile, emotional state, and ideological leanings of specific audiences, using the enormous reservoirs of personal data we have all willingly handed over. So a false story can be subtly adjusted: a slightly different angle for conservative viewers, a different emotional trigger for progressive ones, a different frame for young people, a different fear for older ones. This is industrial-scale manipulation. The Philippine Nobel Peace Laureate Maria Ressa, whom the author interviewed three years ago, predicted this storm. She showed him the attacks she had received on social media: hundreds of thousands of insults, humiliations, and sexual photomontages, all signed by anonymous users. The International Center for Journalists estimated the number of messages at over 500,000. What is striking, she said, is that someone would invest so much money and effort to attack her. But the point is even darker: what used to require time, technical expertise, and significant resources to manufacture such a campaign can now be done overnight, automatically, and at almost no cost. We have already seen fake images circulate as breaking news before anyone could verify them. We have heard manipulated audio presented as authentic statements from politicians. We have watched content specifically fabricated to inflame a particular election, a protest, or a war. The problem is no longer simply distinguishing the true from the false. The problem is that the false can be produced far more quickly than society can possibly verify it. It’s a race where one side is sprinting and the other is still trying to tie its shoelaces.
This is not just a technical problem; it is a crisis for journalism, and for everyone who still believes in the old-fashioned idea that facts matter. Over four years, the author held long conversations with some 140 journalists from different countries, from wealthy democracies to war zones, from elite global newsrooms to independent freelancers working alone. They were a diverse group, but their accounts formed a surprisingly consistent X-ray of the profession at a critical moment. The great challenge, many of them argued, is no longer simply reporting what happens. It’s making it matter. Because the truth, by itself, no longer prevails. Think about that for a moment. In the old media ecosystem, newsrooms had enormous power to decide what was relevant, to cross-check facts, to provide context, and to build a common public conversation. They were imperfect, often deeply so, but they served as a kind of referee. That position has been swept away by the digital platforms that now govern how people consume information. Multinational technology companies have changed the rules of distribution, and with them, the relationship between what is important and what is visible. Now, the vital competes directly with the viral. A carefully reported investigation into corruption sits side by side with a ridiculous conspiracy video, as if they were equally deserving of your attention. An explanation competes with a reaction; a nuanced analysis competes with a screaming headline; a journalist’s professional judgment competes with an algorithm that doesn’t know or care what is true. And in that brutal competition for attention, the odds are stacked against accuracy. Platforms are not designed to distinguish truth from falsehood; they are designed to maximize engagement, and engagement tends to flow toward emotion, simplification, conflict, and outrage. Algorithms, built with commercial logic, reward the most striking content regardless of whether it is true, misleading, or deliberately manipulated. The result is an ecosystem in which information loses density, public debate becomes polarised, and trust in institutions crumbles. It is not an accident; it’s the architecture. As absurd as it sounds, we now live in a world where populism is not just a social or political movement, but a kind of algorithmic outcome, a feedback loop between the emotional states of a population and the content that is fed to them. Some leaders are born, or at least sustained, by their ability to read and exploit the data-driven moods of the crowd.
This is what makes misinformation so insidious. It has ceased to be a malfunction of the digital system; it has become part of its core logic. It is no longer just a tool used by a few bad actors; it is an architecture designed to influence, confuse, and erode societies from within. It doesn’t confront the truth directly, because direct confrontation can be exposed and defeated. Instead, it dilutes the truth, buries it in noise, and makes people feel that there is no such thing as truth anyway. That is far more dangerous and effective. When you can make people believe that every source is untrustworthy, you don’t need to convince them of a lie; you just need to prevent them from agreeing on any reality at all. And once society can no longer agree on basic facts, it can no longer function as a democracy. It cannot hold leaders accountable, because every criticism is dismissed as fake news. It cannot address common problems, like climate change or public health, because every expert is accused of bias. It cannot even have a civil conversation, because every disagreement becomes a war over reality itself. This is not a hypothetical future. We are already living in it. Every time a manipulated video goes viral, every time a baseless rumour moves a stock market, every time a false story influences an election, the shared foundation of our society cracks a little more. And the responsibility for this cannot be laid solely at the feet of foreign trolls or evil hackers. We are all participants. We share the content that makes us angry. We click on the headline that confirms our suspicions. We trust our favourite influencer more than the scientist or the reporter. And the algorithms learn from us, adapting to our weaknesses, giving us more of what we already want to believe. The enemy is not a distant foreign power; the enemy is our own attention, harvested and weaponized at scale.
For journalism, this is a moral test, not just a technological or economic one. Newsrooms have spent the last two decades struggling with a collapsing business model, chasing digital subscriptions, fighting for ad revenue, and experimenting with new formats. But all of that is secondary to the deeper challenge of maintaining standards in an environment that actively pushes them down. It is so tempting to lower the bar, to write for clicks, to publish first and verify later, to mirror whatever is trending, to compete on speed rather than substance. But the journalists who spoke for this article insist that the opposite is necessary. Journalism must impose criteria over acceleration. It must protect rigor when the algorithms reward speed and outrage. It must continue to play on the field of credibility, not attention. This does not mean rejecting AI; quite the opposite. AI can become an extraordinary tool for journalism if it is used to boost quality rather than just speed or volume. Already, smart newsrooms are using machine learning to sift through enormous datasets, to detect patterns of corruption, to visualize complex stories in ways that are both beautiful and informative. But the value of journalism will never lie in the tool itself. It will always lie in the original information, in investigation, in judgment, in accountability, in context, in credibility, in the willingness to go to uncomfortable places and ask difficult questions. The camera can be used to manipulate or to document; the pen can be used to deceive or to clarify. The choice remains human. And it is a choice that every journalist, every editor, and every reader makes every single day.
So where does that leave us, ordinary people who are not journalists, not world leaders, not tech engineers? It leaves us with a responsibility that we did not ask for and cannot escape. We live in an age of information abundance and attention scarcity, and the greatest challenge of our time is not technological but ethical: we must decide whether we still value the truth enough to defend it. A society can survive uncertainty, can survive crisis, can survive conflict, can even survive war. But it can hardly survive the disappearance of truth as a shared reference. Without shared facts, there is no shared reality; without shared reality, there is no community, no democracy, no progress. The good news is that the fight is not lost. Awareness is growing. People are learning to pause before they share. They are asking where a piece of information came from, who created it, and why. They are supporting news organizations that do the slow, unglamorous work of verification. They are demanding transparency from platforms instead of accepting their black boxes as inevitable. This article was published in cooperation with World News Day, a global campaign to remind us of the vital role of journalism in providing facts and clarity. The name is a little old-fashioned, but the message is more urgent than ever. We need to start talking about misinformation with the same energy and curiosity we reserve for the latest AI trick. We need to make truth a popular topic of conversation, as engaging as football, as fascinating as the latest chatbot. Because the future of our societies may well depend on whether we can still agree that facts matter, and whether we are willing to do the difficult, unglamorous, constantly recurring work of protecting them. That is not a technical problem; it is a human one. And it is up to us to solve it.