Google's AI Gemini removes watermarks, sparking copyright concerns

Google's new AI model, Gemini, has sparked concerns among creators. The Gemini AI can create images and solve complex problems. However, it can also remove watermarks from licensed images. This feature worries many because it could lead to copyright violations. The Gemini 2 Flash version is specifically being used to erase watermarks with simple text prompts. Many users have reported this issue to Google. They express concerns about how easy it is to misuse the AI for unethical purposes. While Google states that using its AI for such actions breaks its terms of service, critics argue the company needs to strengthen its controls. Other AI firms, like Anthropic and OpenAI, have stricter guidelines to prevent these actions. In response to recent challenges, Google has been revising its AI image tools. Earlier this month, the company apologized for past mistakes and announced updates. It also introduced the Gemma 3 family of AI models, which improve text and visual reasoning. Since launching in August 2024, the Gemma models have been downloaded over 100 million times and have produced more than 60,000 variants.