A state-of-the-art artificial intelligence designed to deter drug use has sparked a furious backlash after critics accused it of doing the exact opposite: turning narcotics into something alluring. The video, produced by a consortium of tech firms and health agencies, uses generative AI to create hyper-realistic depictions of users experiencing euphoric highs. But rather than scaring viewers straight, the imagery has been widely condemned as a dangerous advertisement for illicit substances.
The footage, which went viral on social media within hours of release, shows a young person injecting a substance and then floating through a kaleidoscope of geometric patterns, accompanied by a pulsating ambient soundtrack. The AI model, trained on thousands of hours of psychedelic art and user testimonials, was intended to showcase the brain’s reward system being hijacked. Instead, viewers described the experience as “visually stunning” and “almost magnetic”.
“We have a moral responsibility to ensure our technology doesn’t become a vector for harm,” said Dr. Helena Zhang, a neurologist at University College London who specialises in addiction. “This video is a textbook example of the law of unintended consequences. The AI has no ethical compass, and the humans behind it failed to foresee that a hallucinogenic trip looks, frankly, appealing to a generation raised on music festivals and VR.”
The creators defended the project in a hastily assembled press conference. Marcus Okonkwo, CEO of the startup behind the AI, argued that the video was designed to be “authentic” and that sanitising the drug experience would undermine the message. “If we make it look like a PSAs from the 80s, nobody watches it,” he said. “We wanted to show the raw reality. The problem is, for some people, that reality is a fantasy.”
Social media platforms have been forced to restrict sharing of the video, with some labelling it as “glamorised drug content”. Twitter, Facebook, and TikTok have all issued takedown notices but struggled to keep up with re-uploads. One viral remix, set to a drill beat, has been viewed over 20 million times.
The backlash highlights a growing dilemma for regulators and tech companies alike. As generative AI becomes more sophisticated, the line between persuasion and exploitation blurs. “We’re entering an era where AI can craft experiences that are more addictive than the drugs themselves,” warned Professor Amara Singh, a digital ethics researcher at Oxford. “This video is a shot across the bow. We need guardrails before someone uses this tech to directly market narcotics.”
In response, the UK Home Office has launched an investigation into the use of AI in public health campaigns. A spokesperson said: “We must ensure that all materials designed to protect the public do not inadvertently cause greater harm. We will be working with experts to establish clear guidelines for the use of generative AI in sensitive contexts.”
The incident has also reignited debates about digital sovereignty. Critics argue that the video was exported globally without cultural context, leading to a one-size-fits-all approach that ignores local sensibilities. “What may be shocking in London is a Tuesday night in Berlin,” said Dr. Zhang. “We cannot outsource our moral judgments to an algorithm.”
For now, the video remains a cautionary tale. As one commentator put it: “We asked AI to show us hell, and it painted us a picture of paradise. That’s the real nightmare.”
