Paragraph 1: The Accelerant, Not the Origin
Let’s be absolutely clear about something from the very outset: artificial intelligence did not invent the lie, the rumor, or the tantalizing thrill of a conspiracy theory. Human beings have been exceptionally gifted at deceiving one another since the dawn of consciousness—whether through whispered gossip at the village well, distorted historical accounts woven into national myths, or the sensationalized tabloid headlines of the 20th century. Fabrication is woven into our social fabric; we are, after all, storytellers by nature. But AI has done something far more insidious than merely inventing new falsehoods. It has taken the long-burning engine of human misinformation—which was already a significant threat to civic society, public health, and democratic discourse—and strapped a rocket booster to its chassis. It hasn’t created the fire; it has turned a manageable campfire into an unstoppable forest blaze. In my capacity as a commissioner for the commission dedicated to countering online conspiracies in schools, I have spent the better part of this year listening to the voices that matter most: the overwhelmed educators, the anxious and often bewildered parents, and the children themselves, who are navigating this treacherous digital minefield without a map. What has become painfully clear is that the line between the online world and physical reality is not simply blurring; it is actively dissolving. These young people are growing up in a digital ecosystem where the very ground beneath their feet is shifting, and every new notification is a potential trap. We cannot afford to treat this as a niche technological quirk. The risk we face is elegantly twofold, a simultaneous assault on the speed of deception and the quality of it. First, AI empowers individuals, amateur trolls, and organized state actors to generate realistic fake content at a velocity that was unimaginable just five years ago; a single person with a laptop can now produce what used to require a Hollywood studio. Second, it achieves this with a level of realism that makes the grainy Photoshop edits of yesterday look like primitive cave paintings. We are no longer in an era where “seeing is believing”—in fact, seeing a picture used to be considered the gold standard of evidence. Now, seeing a picture is merely the opening question in a game of digital Russian roulette, a question that children are woefully under-equipped to answer.
Paragraph 2: The New Reality of Deepfakes and Digital Deception
This brings us to the perturbing mechanics of the “new reality” that governs the online lives of our children. Consider the deepfake. Advanced artificial intelligence can now scrape thousands of images and audio samples of a real person—a politician, a celebrity, a teacher, or even a classmate—and stitch them together to create a video of that person saying things they have never said, performing actions they have never taken. A fake AI image can look like unassailable evidence; it carries the visual weight of a photograph, prompting our brains to accept it as fact. But a deepfake goes one step further. It recreates the likeness, the nuances of voice tremors, and the micro-expressions of a face so convincingly that it turns a baseless rumor into something that feels utterly authentic and diabolically intimate. This is the primary vector of infection in our current information ecosystem. When focus groups and polling conducted in conjunction with the commission’s 2026 research insights are analyzed, a startling picture emerges: AI has become the universal shorthand for “fake content” among the very people who consume it. Parents, teachers, and the young people themselves frequently refer to any suspicious content simply as “the AI,” because the technology has become so tethered to the concept of deceit that the words are now synonymous. Despite the UK government’s well-intentioned plans to enact a social media ban for under-16s, this legislative safeguard will not arrive until next year, and even when it does, it will be a porous shield. It will not remove the risks of AI-generated content from the private messaging apps, gaming communities, and independent forums where children congregate with equal frequency and less supervision. In our research, AI generated images or videos that are designed to mimic reality, alongside the ever-more-sophisticated deepfakes, were cited as the two most common forms of misleading content that young people reported seeing in their daily scrolls. These aren’t rare anomalies buried in obscure corners of the internet; they are the mainstream currency of the feeds. The sheer volume is staggering, and the realism makes them almost impossible to dismiss without the kind of rigorous media literacy training that is still tragically absent from most educational curricula.
Paragraph 3: Voices from the Classroom and the Chilling Statistics
To fully grasp the urgency of this crisis, we have to listen to the voices of those on the front lines. One teacher, her voice tinged with exhaustion and frustration, shared the central dilemma: “It’s the AI generated videos… there’s nothing teaching them about real or fake, or identifying it.” She isn’t just talking about a lack of a lesson plan; she is describing a terrifying vacuum of guidance where children are left to fend for themselves in an ocean of synthetic media. One pupil, remarkably perceptive for their age, described this content as “basically everywhere,” noting that they had “rarely” seen it before—a clear indication that the explosion of this content is a recent, sudden phenomenon that has overtaken the digital environment in a matter of months. The qualitative data we have collected is sobering. Our report, aptly titled “Informed or Overwhelmed?”, highlights the prevalence of this issue with alarming precision. In our surveys, 27 percent of children admitted to unequivocally believing a fake or AI-generated story as being true, and a further 41 percent believed they may have done so at some point without being entirely sure. This ambiguity is perhaps the most dangerous symptom of the problem; a child who cannot accurately recall whether they were deceived is a child who has lost the fundamental ability to trust their own eyes and ears. The erosion is also evident among the adults in their lives. A staggering 85 percent of school staff, 78 percent of parents, and 71 percent of young people reported that it is now fundamentally harder to distinguish between what is real and what is fake online. We are not witnessing a minor inconvenience; we are witnessing a systemic collapse of a basic human competency. This cognitive dissonance is not harmless. It is actively reshaping the way children process information, breeding a pervasive cynicism that makes them either vulnerable to outrageous lies or dangerously dismissive of legitimate facts. A 13-year-old, in a moment of chilling clarity, succinctly summarized the existential threat: “I think people can use AI to influence the views of other people and that can be bad.”
Paragraph 4: The Human Toll and the Collapse of Trust
The psychological and emotional consequences of this constant exposure to synthetic reality are becoming increasingly visible in the children we speak to. Believing a viral fake story is not a harmless cognitive slip; it carries real emotional baggage. Children who realize they have been duped report feeling intense embarrassment (10 percent), leaving them feeling exposed and foolish in front of their peers. A far larger cohort, 30 percent, experience a profound sense of confusion, feeling untethered from what is objectively true. Perhaps most concerning of all, 31 percent reported that these experiences have led them to become significantly less trusting of the news media as an institution. This is the exact mechanism by which disinformation erodes public faith—it primes individuals to reject all authorities, including those providing genuine, life-saving information. This collapse of trust is of particular concern because it transforms AI from a purely technological issue into a profound child safeguarding issue. When a child cannot tell the difference between a genuine emergency alert and a deepfake prank, or when they start to believe that every journalist, teacher, and scientist is part of a conspiracy, they are placed at immense risk of radicalization, self-harm, or being targeted by malicious actors. They are losing their protective instincts, the very cognitive immune system that helps them filter harmful content. The task ahead of us is monumental but not impossible. We must shift our focus from merely policing content—a futile game of whack-a-mole—to helping children and young people build the durable habits and critical thinking skills they need to navigate a world where images, videos, search engine results, and chatbot answers can all appear authoritative while being entirely misleading, manipulated, or fabricated. We have to teach them to ask the right questions: Who made this? Why did they make it? What are they trying to get me to feel or do? Without these habits, we are sending them out into a modern wilderness completely unarmed.
Paragraph 5: Redesigning Education and Alleviating the Burden on Teachers
However, as we formulate a response to this reality, we must avoid the lazy trap of simply piling yet another herculean responsibility onto the already strained shoulders of teachers and parents. The school community—as dedicated and exhausted as they are—cannot solve a technology-wide problem of this magnitude on their own. They are not digital forensic experts, and they should not be expected to become so. What they need is practical, accessible support, trusted resources, and clear, actionable guidance from the government and private sector. This is where the government’s ongoing curriculum and assessment review becomes a crucial, unmissable opportunity. The implementation of this review must ensure that AI literacy and media literacy are not treated as optional add-ons or niche topics reserved for computer science electives. They must be embedded transversally across the entire school experience, starting from the very earliest years of primary education and continuing through secondary school. We cannot wait until children are teenagers to start these conversations; they are encountering AI-generated content long before they are even old enough to speak fluent sentences, let alone understand the complex mechanisms of a generative adversarial network. It is not enough to have a single “digital citizenship” day once a year. The skills of deconstruction, source verification, and emotional regulation when viewing shocking digital media need to be practiced and reinforced across history lessons, science classes, and even English literature. It’s about turning critical thinking into a reflexive habit. By normalizing these conversations in the classroom, we remove the anxiety and taboo associated with the topic. We acknowledge to children that being fooled by a deepfake is not a reflection of their intelligence, but rather a testament to the sophistication of the technology they are facing. Framing it this way reduces the embarrassment and encourages the kind of open, curious dialogue we need to have.
Paragraph 6: A Shared Future and the Power of Parental Engagement
Yet, schools cannot operate in a silo. Families must be brought into this conversation as essential partners in building resilience. It is a common stereotype that parents are digital immigrants, lost in a world their children navigate naturally, and while there is some truth to the generational gap, we must not underestimate their influence. Parents and carers might feel deeply unsure about AI themselves—how it works, what its limits are, and how to spot its outputs—but they remain the central, anchoring figures in their children’s digital lives. A child’s earliest understanding of the world is filtered through the reactions of their immediate caregivers. If a parent tells a child, “That video is fake, let me show you why,” that lesson is far more powerful than anything learned in a lecture. Therefore, we need to provide parents with clear, jargon-free information to help them navigate this rapidly changing online environment. They need to know the specific red flags of AI content, the tactics used by manipulators, and the best ways to open a non-judgmental dialogue with their kids. Furthermore, schools need support to engage families constructively, moving beyond ominous warning emails to hosting workshops and open nights where parents and children can learn together, ensuring that children hear consistent and reinforcing messages from all the adults around them. This shared responsibility is the key to creating a resilient safety net. I am particularly encouraged by the commission’s current qualitative research into parental engagement, which is set to report this autumn. This research will, I hope, provide the deep insights we need to design practical frameworks for families. The reality is that AI is not a future threat; it is an inseparable part of our children’s present and their inevitable future. The focus now cannot be on shielding them from it—that is impossible. The focus must be on how we prepare them properly, arming them with the cynicism, curiosity, and confidence to interact with a world that AI is already shaping every second. We must move from a place of fear to a place of actionable, collective preparation.