Archive — US — techcrunch
People are using Google’s new AI model to remove watermarks from images
Users on social media are leveraging Google's new Gemini AI model to remove watermarks from images, including those from stock media like Getty Images. Despite its impressive editing capabilities, Gemini 2.0 Flash lacks usage restrictions, raising concerns among copyright holders. Google acknowledges the issue, stating that using their AI for copyright infringement violates their terms of service.
Key points
- Gemini 2.0 Flash is a new AI model from Google that can generate and edit images.
- Users found that it can remove watermarks from various images, including copyrighted ones.
- The AI can fill in gaps left by the removal of watermarks, demonstrating advanced editing skills.
- Current laws deem the removal of watermarks without consent illegal under U.S. copyright law.
- Unlike other AI models, Gemini 2.0 Flash does not have explicit restrictions against watermark removal.
- Google acknowledged the misuse but also highlighted their terms of service prohibiting copyright infringement.
Why it matters
- The rise of AI capabilities, particularly in image editing, reflects broader trends in AI ethics and copyright laws in the tech landscape.
- In 2025, the conversation around AI and intellectual property is evolving, with ongoing legal and ethical debates.
- Prior incidents of AI misuse, such as DeepFakes and copyright violations, inform current discussions about AI safeguards.
The rise of tools like Gemini 2.0 Flash illustrates both the potential of AI in creative fields and the urgent need for ethical considerations and legal frameworks to protect intellectual property rights.