The above video might not be the one referred to by the Times but not problem as there are many of them and as you can see they are very realistic. This one has been viewed 23 million times. How many would have believed it was genuine but for the labelling on the video. Rough guess: at least 10 million. Think what is being done as you read this by malign actors wishing to cause mischief by misinforming the public.
The problem is many people simply don’t care if the video they are watching is fake or real. If it entertains that is all that matters. AI can therefore get away with this weakness. You wonder where it will all end up. European societies are already under stress with the potential for real societal issues developing.
A recent item in The Times described a video circulating online in which a tiger slips into a quiet suburban garden, stalks a sleeping dog, and is dramatically driven off by a courageous domestic cat. The clip is grainy, shaky and seemingly authentic. It resembles the millions of rough CCTV and phone-camera moments uploaded every day. Yet the whole sequence was created by artificial intelligence. As is the one below but the AI creation is not mentioned:
This tiger-and-cat micro-drama is only one among a growing flood of AI-generated “found footage” videos appearing on YouTube and social media platforms. Many show improbable animal encounters or conveniently heart-warming rescues. Others depict near-miss accidents, implausible crimes, or scenes designed to provoke outrage. They sit at an uneasy crossroads between entertainment, clickbait economics, and the evolving craft of synthetic media.
What makes these videos unsettling is not their content but their plausibility. For decades, most people instinctively trusted visual evidence. Even poor-quality footage carried a presumption of truth. But modern generative tools can fabricate entire events with astonishing realism, complete with shaky-cam artefacts, off-balance framing, and environmental noise designed to mimic amateur recording. The result is a new category of misinformation: videos that feel real to the gut even when the brain has doubts.
This raises three serious concerns.
The first is emotional manipulation. AI-generated clips bypass rational scrutiny because they trigger immediate reactions. A sleeping dog about to be mauled or a miraculous last-minute rescue hits directly at human instinct. These moments spread fast because people share them before they pause to question their authenticity. The architecture of social media rewards speed, not accuracy, creating fertile ground for synthetic stories.
The second concern is narrative flooding. A hostile state or organised group no longer needs a single convincing fake to mislead the public. All that is required is volume. Dozens or hundreds of half-plausible artificial videos, all pushing a particular theme, can drown out verified reporting. Confusion becomes the goal: when people lose confidence in their ability to distinguish real events from fabricated ones, they disengage from reliable sources altogether.
Third, and perhaps most damaging, is the erosion of genuine evidence. As fake videos proliferate, dismissing real footage becomes easier. Any inconvenient recording can be brushed aside as “AI-generated.” This defensive denial is already visible in political conflict zones, where authentic documentation is routinely labelled fabricated.
There are ways to counteract the problem. Camera manufacturers and smartphone makers are exploring cryptographic watermarking that can prove when footage is genuine. Platforms can label synthetic content clearly without treating it as inherently suspect. Legislators can create penalties for those who deliberately distribute harmful fabrications. Most importantly, the public needs new habits of scepticism: treating online video as raw information that may require verification.
AI-generated entertainment will not disappear, and not all such videos are harmful. But the tiger stalking a garden should remind us of a new reality. In an age where a convincing video can be conjured from nothing, trust itself becomes a resource worth protecting.
