The Detection Game is an interactive online experience where users test their ability to identify AI-generated art against an automated detector. Players classify 10 images as either human-made or AI-generated, then see their results on a leaderboard comparing human vs AI detector accuracy.
This game invites audiences to think critically about AI-generated content and consider ethical ways to use this technology without harming creative communities. It questions our ability to distinguish real from synthetic and serves as a reminder of the importance of media literacy in our digital age.
I developed the AI detector by training a DINOv3 vision transformer on over 185,000 art images across 10 styles, achieving robust performance even on unseen generators like Midjourney, DALL-E, and DreamStudio.
As a visual artist and designer who has witnessed AI's rapid impact on creative fields, this work explores the urgent need for detection methods to preserve artistic integrity, copyrights, and foster dialogue about transparency, authorship, and the future of visual culture.
