On the other hand, it stated that “we do not train generative AI on customer content” and that “you own your content.”
Adobe is making revisions to offer more specificity surrounding areas such as artificial intelligence and content ownership, according to a blog post published by the firm. This comes after the company received user uproar regarding its most recent terms of service (ToS). “Your content is yours and will never be used to train any generative AI tool,” wrote Scott Belsky, head of product, and Dana Rao, vice president of law and policy. “Your content is your possession.”
Infuriated by new, ambiguous phrasing that many took to suggest that Adobe may freely utilize their work to train the company’s generative AI models, subscribers who use Adobe products such as Photoshop, Premiere Pro, and Lightroom were outraged. To put it another way, creators believed that Adobe may utilize artificial intelligence to steal their work and then resell it.
It was believed that other language meant that the firm could truly assume possession of the copyrighted material that users had created (which is understandable when you see it).
According to Adobe, none of that was true. The company also mentioned that the new terms of use were implemented for its product improvement program and content moderation for legal concerns, primarily concerning CSAM. On the other hand, a significant number of customers did not see it in that manner, and Belsky agreed that the corporation “could have been clearer” with the revised Terms of Service.
“In a world where customers are anxious about how their data is used, and how generative AI models are trained, it is the responsibility of companies that host customer data and content to declare their policies not just publicly, but in their legally binding Terms of Use,” Belsky said in his presentation.
In order to achieve this goal, the firm stated that it would revise the Terms of Service by use “more plain language and examples to help customers understand what [ToS clauses] mean and why we have them,” as stated in the document.
By providing an update on June 6th, Adobe did not help its own cause by making some small adjustments to the same ambiguous language as the original Terms of Service, and there was no indication that Adobe would apologize for its actions. The fact that subscribers to its Creative Cloud service threatened to resign in large numbers just seemed to add fuel to the flames more than anything else.
Additionally, Adobe asserts that it exclusively uses Adobe Stock photos to train its Firefly system that it has developed. However, numerous artists have remarked that their names being utilized as search phrases in Adobe’s stock footage portal, as Creative Bloq highlighted. The end result is art that is generated by AI and occasionally comes close to imitating the styles of the artists.
Its most recent post is more of a genuine apology, and it provides a comprehensive description of the changes that it intends to make. The firm stressed that consumers have the ability to opt out of its product enhancement programs and that it will more “narrowly tailor” licenses to the activities that are required. This was in addition to the fields of artificial intelligence and copyright. It went on to say that it only examines data that is saved in the cloud and never looks at anything that is stored locally. Last but not least, Adobe has stated that it would be listening to the feedback of customers regarding the new adjustments.