Over the past two decades, we have watched the pillars of public knowledge gradually weaken. John Stuart Mill is probably turning in his grave at this. From social media platforms creating echo chambers and filter bubbles to the flood of user-generated content drowning out expertise to online hostility and ideological policing driving censorship (self-imposed or otherwise), misinformation and disinformation abound. The epistemic principles that sit at the heart of Mill’s influential account of truth-seeking through public reasoning—the free flow of ideas, freedom from the “tyranny of the majority,” and the need for expert voices to rise to the forefront amidst the “noise and clamour of democracy”—are all under significant strain.
We have known about these epistemic “pathologies” for some time now, arising from the modern digital information environment. I am not usually someone who finds herself existentially alarmed by these kinds of technological harms —I generally believe that, once the implications become sufficiently visible and widely understood, societies eventually move to address them—but a recent and otherwise unremarkable event nevertheless triggered a small epistemic panic in my own head. I woke up one morning to find myself paywalled out of the BBC News app as I routinely checked my phone to scan the morning’s news. Whilst still free to those living in the U.K., the BBC has lately moved to charge overseas readers for access.
I did not immediately subscribe to the BBC when confronted with the paywall. Instead, my heart sank as I mentally added up all my existing outgoings: entertainment streaming subscriptions, news subscriptions, podcast subscriptions, music subscriptions, cloud-storage subscriptions, meditation and language-learning subscriptions, and more. Hard decisions have to be made when online entertainment and news publications charge anywhere from seven to seventy-five dollars per month (the Financial Times premium rate) for access.
There is nothing especially new about paywalls, but the context in which I encountered this one was important. I had recently had a frustrating conversation with a friend whose conspiracy-mindset worldview is accompanied by a highly selective belief in scientific and other widely verifiable facts. In trying to challenge some of their reasoning, I had found myself repeatedly urging them to read more broadly, to compare how reputable news organizations across different countries were reporting on the same events, rather than relying on whatever Apple News happened to surface algorithmically. It struck me that morning that the ability to casually cross-reference and verify information in this way may not remain as readily available as we have long assumed. Around the same time, I had also read a recently published paper by Jan Kulveit, Raymond Douglas, Nora Ammann, et al. titled “Gradual Disempowerment: Systemic Existential Risks from Incremental AI Development.” The paper essentially argues that, in this era of growing technosocial complexity—where technological, political, economic, and social systems have become deeply intertwined—existential disempowerment may not arrive as the dramatic AI takeover we often imagine, but instead it may arrive through the gradual accumulation of individually reasonable commercial or political decisions that, over time, produce serious damage at the societal level. With all of this in my head, the incidental paywall block became particularly poignant.
Online news publication paywalls are a perfect example of what Kulveit et al. describe: rational and justifiable decisions made within commercial environments that do not seem existentially alarming in isolation but which, when situated within the wider information ecosystem and considered alongside what we already know about epistemic processes, begin to reveal the potential for genuine societal damage. Of course, it is rational and reasonable that the BBC should charge people who do not live in the U.K. for access to its news content or that any other reputable news publication should charge for access. Financial independence in the media is critical for information to remain trustworthy and free from external influence but, like any other company, news organizations face competitive pressures and need to survive. Good journalism is expensive. News organization must pay for trained reporters, editors, legal review, extensive verification processes, and travel expenses. And, arguably, the more reputable the publisher, the higher the production costs.
However, though news paywalls are commercially reasonable, this is where Kulveit et al.’s broader systems picture starts to matter—because the economic pressures on journalism are not unfolding in a vacuum. They are emerging within a wider information environment already transformed by algorithmic distribution systems, engagement optimization, changing patterns of media consumption, declining trust, and the growing dominance of online personalities as sources of news and opinion. When viewed within that broader ecosystem, the consequences of restricting access to trusted verification sources begin to look potentially far more serious.
Statistics from the latest Digital News Report from Reuters and the University of Oxford help illustrate how these trends are converging. In 2025, social media and video platforms overtook both television news and news websites for the first time in the United States as the public’s primary source of news. Reuters describes this as part of a broader shift away from traditional, edited journalism and towards online personalities, influencers, and creators.
This shift reveals a deeper contradiction in how people now consume information. Although online creators and personalities are increasingly becoming many people’s first source for news, respondents also identified them as one of the two biggest sources of misleading or false information online, tied with government propaganda. Most respondents said they worry not only about the spread of misinformation itself but about their own ability to distinguish fact from falsehood more generally. And whilst people may increasingly discover information through social media and online personalities, when they encounter something they suspect may be misleading, they still overwhelmingly turn to established and trusted news brands to fact-check it.
To add to this contradiction, while nine out of ten households in the U.S. subscribe to video-streaming services, only two in ten subscribe to news services. But, as I said with respect to my own situation, decisions have to be made, and, despite the apparently high incomes in the U.S., a substantial proportion of Americans have relatively little truly discretionary income once housing, healthcare, transport, and insurance are deducted. And, increasingly, the information people rely upon to verify what is true competes economically with entertainment services designed to capture attention.
What this research tells us is that societies (at least in the U.S.) are becoming more dependent on platforms and personalities they trust less, while the institutions they rely upon for verification are becoming less central to everyday information consumption. Adding paywalls would make them also much less accessible.
The problem is that the profound transformation of our information ecosystem is already making it increasingly difficult to maintain the epistemic conditions that allow the general public to reason soundly about the world. The effects of this transformation have already placed enormous strain on public knowledge. When we look at the information ecosystem as a whole and consider how separate, often reasonable, decisions and processes can collectively reshape it, adding paywalls to an already fraught information environment may have deeper consequences than we realize. In a space already distorted by algorithmically amplified misinformation and weakened epistemic friction, the danger is not simply that society will come to know less, but that it may gradually lose the practical habits and pathways through which knowledge itself is established.
Alexandra Frye
Alexandra Frye edits the Tech & Society series, where she brings philosophy into conversations about tech and AI. With a background in advertising and a master’s in philosophy focused on tech ethics, she now works as a responsible AI consultant and advocate.
