Viral AI images teach people how to spot AI.
For over a century, photography has been one of journalism’s most powerful tools in establishing truth and trust. Then AI image generation took the world by storm.
What started as a revolutionary creative tool quickly blurred the line between real and fake, threatening the credibility of journalism. So, working with the leading voice in press criticism, Columbia Journalism Review, we decided to take action - by turning AI against itself.
We launched The PSAi - a music video where the viral AI images spreading misinformation taught the world how to spot AI.
The Music Video
We started by scouring the internet for AI-generated images that had already reached millions. Then, using AI against itself, we employed generative video models to animate these static images, and layered in human performance through facial mapping to lip-sync the characters.
To carry the message, we wrote an original song where every verse included specific tips and visual cues to help people spot AI red flags. But we knew a dry PSA wouldn’t stick. So instead of a somber tone, we leaned into a playful, 90s-inspired rap style, designed to grab attention while delivering something genuinely useful.
The Website
People who saw the music video were driven to the campaign website, where they could learn more about each of the viral images, and get further details on how to spot AI-generated imagery.
Getting Into Schools
The PSAi even became part of the elementary school curriculum, teaching the next generation an important new behavior: question everything you see before you share.
