Artists take legal action against Google: copyright battle over Imagen AI training

Visual artists have initiated a lawsuit against Google and Alphabet, alleging copyright infringement through the unauthorized use of their works to train Imagen AI, raising a broader debate on the intersection of technology and intellectual property rights.
Person, Security

A legal battle is unfolding as visual artists take on Google and its parent company Alphabet, over the use of their copyrighted works without consent to train their AI image generator, Imagen. The case, brought forth by prominent artists including Jingna Zhang, Sarah Andersen, Hope Larson, and Jessica Fink, alleges that Google’s Imagen AI used their copyrighted artwork as part of its training data without permission. This lawsuit highlights a significant copyright issue for AI development, where training datasets often contain millions of images scraped from various sources, potentially including copyrighted material.

The artists are seeking legal remedies, including damages and the destruction of all copies of their works allegedly made during Imagen’s training. This legal action echoes a broader concern among creative professionals about the implications of AI on intellectual property rights. 

The case’s progression through the judicial system is being closely watched, as it may set significant precedents for future AI copyright disputes and comes as one of many similar AI copyright infringement lawsuits:

  • Getty Images vs. Stability AI: Getty Images sued Stability AI in both the US and UK for using millions of its copyrighted images to train the AI model Stable Diffusion without permission. 
  • Visual Artists vs. Stability AI, Midjourney, and DeviantArt: A class action by visual artists against AI developers for using copyrighted artworks to train AI models without consent. The artists argue that the AI-generated images are derivative works, raising questions about the boundaries of originality and reproduction in AI outputs.
  • Programmers vs. GitHub and OpenAI: Programmers filed a lawsuit against GitHub, Microsoft, and OpenAI alleging that their code was used without proper licensing to train AI models.

The U.S. Copyright Office has actively sought public input on how to handle copyright infringements involving AI. This includes discussions on whether there’s enough human involvement in AI outputs to consider them original creations by humans or if they are mere reproductions of scraped data .

Major tech firms argue that their use of copyrighted content in AI training is a form of “fair use” aimed at identifying patterns rather than replicating the original works. This stance is part of a broader industry argument that AI’s training processes do not specifically aim to extract or reproduce individual copyrighted works but rather to learn from a broad corpus of data to generate new content. 

As the legal battles unfold, the outcomes will likely influence how copyright laws are interpreted in the context of AI. The discussions around these cases reflect a growing recognition of the need for updated legal frameworks that adequately address the nuances of AI technology and its impact on creative industries.

This ongoing situation presents a critical juncture for the intersection of technology and copyright law, offering an opportunity to redefine the boundaries of authorship and intellectual property in the digital context.