It seems like AI generated content is all over social media these days. Sometimes I see funny cats on Instagram and laugh and then I realize the cats aren't real. That's pretty sad.
AIgatha Christie offers a solution. With a pipeline that checks for digital watermarks and uses state of the art AI image detection, it identifies AI images that are too good for humans to pick up on. It also integrates human feedback, allowing communities to reach a consensus about what material is genuine. Packaged into a Chrome extension, it's easy to quickly verify your hunch about which suspects are AI and which are innocent.
AI track
The Project
AIgatha Christie is a chrome extension that quickly checks if an image is AI generated or not. It uses a pipeline including C2PA provenance and the Sightengine API.
The Team
Will Soylemez: Worked on the backend API
Gabe Klaris: Made the Chrome extension
Acknowledgements
Uses the Sightengine library as mentioned. A lot of the code was written by Claude Code.
Reflection
We chose this project because verifying the legitimacy of images online is increasingly difficult. It's frustrating not knowing whether the media you view is real or not. Although existing tools aren't perfect, having an easily-accessible pipeline like this is valuable because it can be expanded on as technology improves and best practices around content veracity evolve.
Built With
- c2pa
- chrome
- express.js
- sightengine
Log in or sign up for Devpost to join the conversation.