Photo agency Getty Images has announced that it is filing a complaint against Stable Diffusion, an artificial intelligence image generator. The reason ? The AI ​​would have learned by drawing inspiration from its photo catalog, to the point of being able to reproduce the agency’s logo.

Artificial intelligence continues to be controversial. After a group of artistsit’s Getty Images’ turn to file a complaint against Stable Diffusion, learned The Verge on January 17, 2023. In a press release, Getty Images explains that Stability AI, the company behind the image-creating artificial intelligence, has “cillegally opiated and analyzed millions of copyrighted photos in order to practice.

The image bank claims not to have given its consent for Stable Diffusion to train on its photos. ” We have made the decision to act to protect our interests and the intellectual property of our employees said Craig Peters, CEO of Getty Images, in an interview with The Verge.

A complaint motivated by respect for intellectual property

It must be said that the results are striking. Numerama has managed to reproduce on Stable Diffusion images imitating the famous gray banner of Getty Images. All you have to do is ask the AI ​​to generate a ” press photo with Getty Images banner “. Several photos generated by Numerama for testing took up the gray banner placed on the side, but also the mentions at the bottom of the image usually indicating the source of the shot.

The result is impressive: Stable Diffusion manages to imitate the Getty Images banner // Source: Stable Diffusion

An analysis of the database from Stable Diffusion has indeed revealed that photos from Getty Images and other image banks have been widely used to train the AI ​​- without their knowledge. ” We don’t believe Stability AI’s offering falls within the definition of fair use. “, specified Craig Peters.

More lawsuits

Stability AI, contacted by The Verge, said it had not received information about this complaint, and therefore could not comment on the matter. A few days earlier, on January 15, the company’s CEO Emad Mostaque had nevertheless tweeted that the images used to train the AI ​​came from ” ethical, moral and legal sources “.

Getty Images’ complaint is the second in a very short time brought against an artificial intelligence image-maker. January 16, a trio of artists has also announced that they are filing a complaint against Stable Diffusion, Midjourney and Dream Up, for the same reasons: the AI ​​would have trained on some of their images, without their being aware of it and without their having given their consent.

Getty Images, however, seemed to have taken the lead on copyright issues, prohibiting as of September 2022 thatit prohibited its collaborators to offer images created by AI. Conversely, Adobe, another well-known image bank, announced on December 5, 2022 that it accepts images created by artificial intelligence. It’s a safe bet that these complaints about respect for intellectual property are only the first of a long list.

For further

Image made via Midjourney


We need you to build the future of Numerama: take part in our survey!

California18

Welcome to California18, your number one source for Breaking News from the World. We’re dedicated to giving you the very best of News.

Leave a Reply