Adobe Inc. plans to integrate a full artificial intelligence (AI) image generation tool into its Photoshop software later this year, according to the company's announcement on Tuesday. While Adobe's image and video editing tools are popular among creative professionals, they face increasing competition from startups like OpenAI, Midjourney, and Stability AI, which offer services capable of generating images from text prompts. In an effort to avoid copyright infringement claims, Adobe is developing its own AI system called Firefly, which is trained using data that Adobe has obtained the rights to. Previously, Adobe had already released image-generation tools in Photoshop that enable users to fill in or expand certain parts of an existing image.

At a recent conference in London, the company revealed their upcoming full image generation feature, which will be powered by a new AI system called Firefly Image 3. Adobe's focus has been on enhancing the efficiency of professionals who utilize their software. The new image-generation tool will have the capability to use a user's uploaded image as a reference for generating images with similar compositions. For instance, a designer could quickly sketch a scene on a napkin, take a photo of it using a smartphone, and then instruct Photoshop to generate fully-featured images in various styles, as explained by Ely Greenfield, Adobe's Chief Technology Officer for digital media. According to Greenfield, this feature eliminates the need for meticulous descriptions and ensures that the desired elements are accurately specified. Adobe stated that a test "beta" version of the software is already available to some users, but a specific release date for the general public has not been provided.