In an age when a single prompt can produce a seemingly authoritative article, a convincing health tip, or a fake medical paper within seconds, the question of who separates fact from fiction has become urgent. According to global health experts Kent Buse and Sarah Hawkes of Monash University Malaysia, the answer cannot be left to chance, tech companies, or overburdened individuals. In their timely article, “Who sifts? Knowledge as a public good in the age of AI,” published in PLOS, they argue that sifting truth from AI-generated content should now be regarded as a core responsibility of public health institutions. This is not simply about catching lies or debunking myths. It is about protecting the very foundation of public health: shared, reliable knowledge. When people cannot tell what is true, they cannot make informed decisions about their bodies, their families, or their communities. And when trust in information collapses, so does trust in medicine, science, and government. The authors invite us to think of knowledge not as a commodity to be harvested by algorithms, but as a public good that must be actively tended, defended, and made accessible to everyone. Their argument feels especially human because it begins with a simple, humble admission: in a world saturated with AI-generated noise, none of us can sift alone.
The idea of knowledge as a public good is not new, but it has never been more at risk. In economics, a public good has two defining features: it is non-rivalrous, meaning one person using it does not reduce another’s ability to use it, and non-excludable, meaning no one can be easily kept from benefiting from it. Clean air, national defense, and basic research are classic examples. Health knowledge, too, belongs in this category. Knowing that vaccines save lives, that clean water prevents cholera, and that handwashing reduces infection are pieces of knowledge that protect everyone, regardless of wealth or status. For centuries, public health institutions have played a quiet but vital role in sustaining this knowledge: gathering data, conducting research, issuing warnings, and educating the public. But the digital age, and especially the arrival of generative AI, has altered the landscape. Now a person seeking health advice may encounter a thousand sources, many generated by machines designed to mimic human language, with no underlying understanding of medicine or biology. Some content is deliberately malicious, some is simply sloppy, and some is so convincing that even experts need time to evaluate it. In this chaos, the public good of knowledge is being diluted and polluted. Buse and Hawkes argue that public health institutions must become the sifters—not as gatekeepers in an old-fashioned sense, but as trusted stewards who curate, verify, and clarify the torrent of information. Without this intervention, the very concept of shared truth begins to crumble, and with it, the social solidarity that makes public health possible.
The question “Who sifts?” is deceptively simple. In practice, it raises difficult issues of responsibility, power, and capacity. One tempting answer is that individuals should sift for themselves, developing critical thinking skills and media literacy. But this is both unfair and unrealistic. The average person, especially during a health crisis, is already worried, busy, and confronted with rapidly evolving information. Asking them to evaluate the technical details of a vaccine trial or to judge whether a viral video of a doctor is authentic is too much. Moreover, AI has made misinformation far more scalable and customizable. It can adapt to a person’s language, cultural context, and emotional triggers, making it nearly impossible to resist without institutional support. Another tempting answer is that technology platforms should sift through their own content. Yet social media companies are not primarily designed to promote truth; they are designed to promote engagement, and false, sensational, or inflammatory content often generates more engagement than accurate, measured information. Even if companies attempt to label AI-generated content, enforcement is inconsistent and often reactive. Buse and Hawkes propose a different path: public health institutions themselves must take on the role of “sifter-in-chief.” This means creating dedicated teams to monitor AI-generated health content, partnering with researchers to develop detection tools, and publishing clear, accessible guidance that communities can trust. It also means recognizing that sifting is not a one-time correction but an ongoing process, because AI continues to generate and spread content around the clock. Public health institutions, they argue, already have the mandate, the expertise, and the moral authority to do this work. What they often lack is the resources and the institutional will.
Of course, making sifting a core responsibility is easier said than done. Public health institutions across the world are frequently underfunded and overstretched, especially after years of responding to COVID-19. They are asked to do more with less, and now a new mandate could feel like an impossible burden. There are also serious risks. Attempts to debunk misinformation can sometimes amplify it, a phenomenon known as the “backfire effect,” where people dig in deeper when they feel attacked. Institutional attempts to sift truth might also be perceived as censorship, especially in polarised societies where trust in experts is low. Buse and Hawkes are not naive about these challenges. They call for a collaborative, transparent approach, where public health institutions do not act as solitary judges but as conveners of discussion. They emphasize the importance of working with grassroots community organisations, educators, journalists, and tech hubs, and of understanding that sifting must be culturally sensitive. A fact sheet that works in one community may not resonate in another; a meme that is true may still be harmful if shared without nuance. The authors also stress the need for investment in the human dimensions of sifting: training communicators who can translate complex science into ordinary language, building hotlines and chat services where people can ask questions, and creating spaces for honest dialogue about uncertainty. In this sense, sifting is not merely an intellectual exercise; it is an act of care. It means meeting people where they are, listening to their fears, and providing guidance that respects their lived experiences. Public health institutions that embrace this role will need to be courageous, humble, and adaptive, but the alternative—allowing AI-generated confusion to fill the knowledge vacuum—is far more dangerous.
The human stakes of this argument are profound. Behind every statistic about misinformation is a real person making a real decision: a mother wondering whether to vaccinate her child, an elderly man deciding whether to follow a home remedy for chest pain, a teenager searching for information about mental health and finding an AI-generated chatbot that validates the darkest thoughts. When public health institutions fail to sift, they abandon these people to a wilderness of competing claims. The consequences are not abstract. Vaccine-preventable diseases re-emerge, treatment is delayed, preventive care is postponed, and health disparities widen. The burden of misinformation falls hardest on those with the least access to reliable resources—low-income communities, ethnic minorities, and people in rural areas. Buse and Hawkes remind us that knowledge is not equally distributed, and AI threatens to make inequality even worse. Wealthy, educated individuals may have the skills and networks to verify what they read, but many people do not. Defining sifting as a core responsibility of public health institutions is, at its heart, an equity issue. It says that every person, regardless of education, income, or language, deserves access to truthful, understandable health information. It also recognizes that trust cannot be bought or imposed; it has to be earned through visible, consistent, and compassionate action. When institutions show up, not as scolds but as allies, they can build relationships that withstand the constant pressure of misinformation. This is the human side of the authors’ message: sifting is not just a technical response to AI. It is a commitment to companionship in an increasingly confusing world.
In their call to action, Buse and Hawkes do not propose abolishing AI or retreating to a pre-digital paradise. Instead, they ask us to imagine a future where public health institutions step into their rightful role as stewards of knowledge. This will require new types of professionals, new forms of collaboration, and above all, new funding. The authors note that their article received no specific funding, which underscores a larger point: sifting truth is currently seen as a nice-to-have rather than a necessity. That must change. Governments, health agencies, and international organisations should allocate money and talent to the creation of public knowledge infrastructure, just as they build hospitals and water systems. They should also encourage research on how AI is shaping health beliefs, not just how to detect fake images. And they should listen to communities about what kind of support they actually need. The phrase “knowledge as a public good” may sound abstract, but it becomes concrete in the simple act of providing a parent with a clear, reliable answer after a sleepless night of searching online. It becomes real when a patient feels comfortable asking a doctor a question because they have learned to trust the institution behind the answer. It becomes real every time a public health official chooses to explain, rather than lecture, and every time a fact-checking team works overtime to stop a viral lie from causing harm. In the age of AI, there is no neutral position. Public health institutions either sift or they let the machines decide. The authors’ message is hopeful: if we treat knowledge as a shared treasure, and if we give our institutions the resources to protect it, we can navigate the AI age without losing ourselves. The task is hard, but the alternative is unthinkable. We owe it to each other, and to the next generations, to take the question “Who sifts?” seriously. Public health is the art and science of protecting people, and there is no better way to protect people than to ensure that what they know is true.