Adobe clarifies it is not using customers’ content for generative AI training

Adobe acknowledged that it should have modernised the language of its terms sooner [File]
| Photo Credit: REUTERS

Adobe has clarified it is not using customers’ content to train generative AI, as users reacted with anger and confusion over the company’s new terms of service, and expressed fears that their work was being scraped for AI training.

In an official post on Monday, Adobe acknowledged that it should have modernised the language of its terms sooner.

The company also explained that it trained its text-to-image generator powering the Adobe Firefly tool on licensed content such as Adobe Stock media, and public domain content with expired copyrights.

“At Adobe, there is no ambiguity in our stance, our commitment to our customers, and innovating responsibly in this space. We’ve never trained generative AI on customer content, taken ownership of a customer’s work, or allowed access to customer content beyond legal requirements. Nor were we considering any of those practices as part of the recent Terms of Use update,” said Adobe in the post, adding that it was still working on its terms of use.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

Adobe’s recently updated Terms of Use panicked many users, who claimed that its ‘Licenses to Your Content’ section was confusing. Many believed it meant Adobe would scrape customers’ copyrighted content to train its generative AI products. Others were angered by Adobe’s content review policies and their implications.

“Your content is yours and will never be used to train any generative AI tool. We will make it clear in the license grant section that any license granted to Adobe to operate its services will not supersede your ownership rights,” said Adobe in its blog post, adding that reviews of uploaded content might take place to make sure customers were not hosting child abuse media.

The company is planning to roll out the updated changes by June 18.