Following months of testing, Adobe on Wednesday made a number of generative AI features available throughout its software. The company has plans to increase prices and pay the contributors whose work made the features possible.
Adobe rolls out AI features: The foundation of Adobe’s subscription software business is its line of editing tools like Photoshop. The business has been gradually integrating new AI features, such the capacity to create images from text, into those programmes during the previous six months.
Content producers are suing technology companies in court to determine whether they owe royalties for the use of their work in “training” AI systems. Meanwhile, Adobe is reassuring businesses that the content generated by its systems will be legally safe to use. In order to support its claims, the company is providing its clients with a financial indemnity. Adobe’s system is built on content that either belongs to it or is in the public domain.
Adobe announced on Wednesday that many of its subscription products will see monthly price increases of $2 to $5 starting in November.
Customers of Adobe will receive set number of “credits” to use generative AI features. Users have the option of purchasing additional credits or continuing to utilise the functions, although at a slower rate.
Adobe also announced that it would compensate authors of the stock picture databases used to hone its AI systems.
“This year, Adobe will award artists one-time ‘contributor’ bonus based on the number of photos they contributed to Adobe’s database and the number of times conventional methods licensed their photographs from June 3, 2022, to June 3, 2023.”
Following that, Adobe will start awarding the bonus on an annual basis for the training work performed using its AI systems.
According to Ely Greenfield, chief technology officer for digital media at Adobe, “We want our stock contributors to continue contributing both for the stock market, which is paying out more than it ever has, and for the value they are contributing to the training of these models.”