Developer claims to have cracked Google's SynthID watermarking system using just 200 Gemini images and basic signal processing, though Google disputes the claim.

A developer named Aloshdenny claims to have reverse-engineered Google DeepMind's SynthID watermarking system and open-sourced the work on GitHub

The alleged exploit required only 200 Gemini-generated images, signal processing techniques, and no neural networks or proprietary access

The developer demonstrated how watermarks can be removed from AI-generated images or added to non-AI works, contradicting the system's security premise

Google disputes Aloshdenny's claims, stating the developer's assertions are inaccurate