\n<\/p>\n
Users on social media have discovered a controversial use case for Google\u2019s new Gemini AI model: removing watermarks from images, including from images published by Getty Images and other well-known stock media outfits.<\/p>\n
Last week, Google expanded access to its Gemini 2.0 Flash<\/a> model\u2019s image generation feature, which lets the model natively generate and edit image content. It\u2019s a powerful<\/a> capability<\/a>, by all accounts. But it also appears to have few guardrails. Gemini 2.0 Flash will uncomplainingly create images depicting celebrities<\/a> and copyrighted characters<\/a>, and \u2014 as alluded to earlier \u2014 remove watermarks from existing photos.<\/p>\n New skill unlocked: Gemini 2 Flash model is really awesome at removing watermarks in images! pic.twitter.com\/6QIk0FlfCv<\/a><\/p>\n \u2014 Deedy (@deedydas) March 15, 2025<\/a><\/p>\n<\/blockquote>\n As several X and Reddit<\/a> users noted, Gemini 2.0 Flash won\u2019t just remove watermarks, but attempt to fill in any gaps created by a watermark\u2019s deletion. Other AI-powered tools do this, too, but Gemini 2.0 Flash seems to be exceptionally skilled at it \u2014 and free to use.<\/p>\n Gemini 2.0 Flash, available in Google\u2019s AI studio, is amazing at editing images with simple text prompts. <\/p>\n It also can remove watermarks from images (and puts its own subtle watermark in instead \ud83e\udd23) pic.twitter.com\/ZnHTQJsT1Z<\/a><\/p>\n \u2014 Tanay Jaipuria (@tanayj) March 16, 2025<\/a><\/p>\n<\/blockquote>\n To be clear, Gemini 2.0 Flash\u2019s image generation feature is labeled as \u201cexperimental\u201d and \u201cnot for production use\u201d at the moment, and is only available in Google\u2019s developer-facing tools like AI Studio<\/a>. The model also isn\u2019t a perfect watermark remover. Gemini 2.0 Flash appears to struggle with certain semi-transparent watermarks and watermarks that canvas large portions of images.<\/p>\n Still, some copyright holders will surely take issue with Gemini 2.0 Flash\u2019s lack of usage restrictions. Models including Anthropic\u2019s Claude 3.7 Sonnet<\/a> and OpenAI\u2019s GPT-4o<\/a> explicitly refuse to remove watermarks; Claude calls removing a watermark from an image \u201cunethical and potentially illegal.\u201d<\/p>\n Removing a watermark without the original owner\u2019s consent is considered illegal under U.S. copyright law (according to law firms like this one<\/a>)\u00a0outside of rare exceptions<\/a>.<\/p>\n Google didn\u2019t immediately respond to a request for comment sent outside of normal business hours.<\/p>\n<\/div>\n\n
\n