The speedy progress of AI development continues to dominate headlines, but the latest story to hit the news cycle is more bad press than good press: It turns out that Google's latest Gemini 2.0 Flash image editor is rather adept at automatically removing watermarks from images.

People appear to be using Google's new Gemini model to remove watermarks from images, according to reports on social media.

It's bad enough that tech companies trained their AI models on content that doesn't belong to them. Now, however, it appears that at least one AI model is being used to remove watermarks from images so others can use content that doesn't belong to them,

Tools like Watermark Remover.io can already scrub marks from companies like Shutterstock, and a research team at Google built a watermark removal algorithm in 2017 to highlight the need for more secure protections. Conversely, some AI tools — like OpenAI’s GPT-4o — will refuse requests to remove them.

Gemini 2.0 Flash’s image generation feature is currently only available in developer tools like AI Studio. It also carries the labels “experimental” and “not for production use.” It wouldn’t be too surprising if Google makes some quick changes to the feature to prevent watermark removal.

Tanay Jaipuria, tech analyst and partner at Wing VC, is one of the many Gemini AI 2.0 Flash AI users who have discovered its ability to clear watermarks. The experimental version of Gemini also allows ‘conversational image editing’ when generating pictures from text prompts.

Gemini 2.0 Flash testers found that the AI lets them remove watermarks from copyrighted photos, which is similar to stealing images.

Google's latest AI model, Gemini 2.0 Flash, is celebrated for its capabilities to remove watermarks from images excellently.

Users on social media recently discovered a controversial feature in Google’s Gemini AI model: the ability to remove watermarks from images, including those from major stock image providers like Getty Images.