Shutterstock said it plans to compensate artists whose works were used to train AI image generation models.
Stock image giant Shutterstock is expanding its partnership with OpenAI and will sell AI-generated images on its site.
The collaboration will see OpenAI’s text-to-image generator, DALL-E 2, integrated into Shutterstock “in the coming months.”
DALL-E 2 has attracted a lot of attention this year. When a user inputs a phrase or a string of words into this AI system, they are able to interpret the description and create multiple images based on the text prompt.
Last month, OpenAI removed the waiting list for DALL-E to give wider access to the text-image model and further improve the AI technology.
Shutterstock said its customers will now have direct access to the model’s image-generating capabilities to “enhance their creative workflows.”
Paul Hennessy, CEO of the image company, said the media for expressing creativity is changing. He added that there is a “great responsibility” to ensure that generative technology that drives innovation “is grounded in ethical practices.”
“We have a long history of integrating AI into every part of our business,” Hennessy said. “This expert-level skill makes Shutterstock the perfect partner to help our creative community navigate this new technology.
“We are committed to developing best practices and experiences to achieve our goal of empowering the world to create with confidence.”
The new agreement deepens Shutterstock’s strategic partnership with OpenAI, which began last year.
OpenAI CEO Sam Altman said data licensed from Shutterstock was “critical” to training the DALL-E model.
“We are delighted that Shutterstock is offering DALL-E images to its customers as part of one of the first deployments through our API, and we look forward to future collaborations as artificial intelligence becomes an integral part of workflows. creative work of artists,” added Altman.
Remuneration of artists
AI models such as DALL-E are trained using a massive amount of images. Concerns have been raised that some of these models use copyrighted images as part of their training datasets.
For example, Polish digital artist Greg Rutkowski recently claims that many of his landscape illustrations are used by Stable Diffusion AI to create new images based on his work.
Shutterstock said AI-generated content is the “cumulative effort” of contributing artists. To create new revenue streams for these artists, the company said it plans to provide “additional compensation” to artists whose works have helped develop the AI models.
Shutterstock said it also plans to compensate these contributors through royalties when their intellectual property is used.
Online art communities have been raise issues with the ethics of AI-generated images and began banning them from their sites.
In September, Getty Images also banned AI-generated images due to “open questions” over copyright and uncertainty surrounding the data on which these AI models are trained.
Getty Images CEO Craig Peters said The edge this week that companies “rushing” to commercialize AI images are ignoring the potential legal and ethical issues surrounding the technology.
There are a growing “legal minefield” around using AI-generated images for commercial gain, JumpStory co-founder Jonathan Løw recently told SiliconRepublic.com. He also warned that the legal risk could lie with the end user if their image used for commercial purposes was subject to a copyright dispute.
10 things you need to know straight to your inbox every weekday. Sign up for the brief dailythe summary of essential science and technology news from Silicon Republic.