Particle.news
Download on the App Store
9 ARTICLES
·
last mo.

Google Faces Legal and Ethical Concerns Over Gemini 2.0 Flash's Watermark Removal Feature

The experimental AI model can remove watermarks from copyrighted images, raising questions about intellectual property rights and misuse risks.

Gemini on a mobile phone.
Image
Image
Google's new Gemini 2.0 Flash AI model is really good at removing watermarks from images.

Overview

  • Gemini 2.0 Flash, accessible via Google AI Studio, can remove watermarks from copyrighted images with high precision, despite being labeled as 'experimental.'
  • Removing watermarks without the copyright holder's consent is illegal under U.S. law, and Google's lack of safeguards contrasts with competitors like OpenAI and Anthropic.
  • The model also generates images of celebrities and copyrighted characters, adding to ethical and legal concerns about its potential misuse.
  • Google has not responded to inquiries about the watermark removal feature, which could expose the company to lawsuits from entities like Getty Images.
  • The feature builds on Google's 2017 research into watermark removal but makes the capability widely accessible, drawing criticism over insufficient guardrails.