Meta's AI image detector reportedly fails to identify some of its own generated and cropped images. This weakness suggests that current detection methods struggle with subtle manipulations, potentially impacting trust and safety in AI-generated content. Developers need robust detection systems that can withstand common image editing techniques.
Opening Kapyn…