The 60‑second AI‑generated video, released by Hong Kong’s Correctional Services Department on June 26, aimed to deter drug abuse on the day of the International Day Against Drug Abuse and Illicit Trafficking. The clip uses slick animation, cinematic lighting, and a subtle romantic tone to show people experimenting with cocaine, methamphetamine, heroin, and MDMA, creating an eerily attractive visual character that many viewers found at odds with the message.
Public backlash erupted instantly on social media, with hashtags calling the video “enticing” and “glamorising drugs.” The department is now facing questions about whether the use of AI can create misleading narratives that the audience might mistake for aspirational content instead of preventive warning.
BBC’s Chinese service explains that AI scripts can inadvertently give compounds an aesthetic quality that undermines the deterrent intent. The clip illustrates challenges in algorithm design—specifically how contrast, music, and visual editing can change the perceived moral tone of public safety communications.
For young viewers who consume this content online, the clip raises a broader question on the role of media literacy. It reminds reporters and creators that even the cleanest governmental agendas can be smudged by technology if data sets, designs, and author intent are not aligned with clear ethical lines.
