A government-backed experiment to use artificial intelligence for a youth anti-drug campaign has, in a grimly predictable twist, gone viral for all the wrong reasons. The Home Office, in partnership with a consortium of tech firms, unveiled a synthetic media video last week intended to warn teenagers about the dangers of synthetic opioids. Instead, the AI-generated footage—depicting a hyper-realistic teenager suffering a fatal overdose—has been memed, remixed, and dissected across social media platforms, with many young viewers mocking its overblown dramatics and uncanny valley aesthetics.
Within 48 hours, the hashtag #NotMyLesson had amassed over 200 million views on TikTok alone. The video, produced by a London-based AI studio using generative adversarial networks, was designed to ‘personalise’ the message by morphing the victim’s face into that of the viewer. However, the technology glitched frequently, leaving a chilling but ultimately laughable hybrid visage that users likened to a ‘zombie from a low-budget horror flick’.
Worse still, some minors reported that the mandatory face-scanning step—required to ‘enter’ the video—was used to extract biometric data without explicit consent, raising alarm bells among digital rights groups. The Information Commissioner’s Office has now launched an investigation. This incident is a stark lesson in the perils of deploying AI without understanding its social and ethical context.
The government’s zeal to ‘modernise’ public health messaging has crashed headfirst into the reality that teenagers are digital natives, not naive receptacles. They see through automated propaganda and recoil from algorithmic manipulation. The backlash also underscores a deeper issue: the commodification of our most intimate data under the guise of safety.
Every time a new app or AI tool asks for our face or voice, we must question who holds the keys to that biometric kingdom. The Home Office’s blunder is a textbook case of good intentions paving the road to a digital dystopia. As Silicon Valley expats like me often caution, technology is never neutral.
When you gamify fear and weaponise personal data, the only predictable outcome is a backlash. The real question now is: will policymakers learn from this or double down on more opaque AI schemes? The future of online safety depends on their answer.
