Google’s AI Can Remove Watermarks—And That’s Raising Concerns

Google’s Gemini 2.0 Flash AI model has recently come under scrutiny for its ability to remove watermarks from copyrighted images, including stock photos from companies like Getty Images.

How Does It Work?

Gemini 2.0 Flash isn’t just erasing watermarks—it’s intelligently filling in the gaps left behind, making the edits nearly undetectable. While other AI models, like OpenAI’s GPT-4o and Anthropic’s Claude, refuse to remove watermarks due to ethical and legal concerns, Google’s AI has no such restrictions—and it’s free to use.

Why Is This a Problem?

✅ Watermarks exist to protect copyrighted content ✅ Removing them without permission is illegal under U.S. copyright law ✅ Stock media companies could take legal action against AI-generated infringement

Google’s Response

Google acknowledged the issue, stating that using its AI tools for copyright infringement is against its terms of service. However, since Gemini 2.0 Flash is still in experimental release, Google is monitoring feedback before making further changes.

What’s Next for AI and Copyright?

This controversy raises larger questions about AI’s role in content creation and copyright enforcement. Should AI be allowed to modify copyrighted material without restrictions? And how will companies like Google ensure their models comply with legal and ethical standards?

Check Out the video here.