Adobe Photoshop and Illustrator, two of the most popular image editing tools available, are reshaping to focus on generative AI. As of Wednesday morning, all Adobe Firefly AI tools are available to all subscribers, allowing users to create the deepfakes they’ve always wanted without messing around with finicky filters or brush tools.
Adobe announced Wednesday that a large family of AI applications in its Creative Suite, dubbed Firefly, has exited its six-month beta and is available in the company’s Creative Cloud apps. These include AI Art Generator, a generative fill tool that can artificially expand image boundaries by creating content outside the frame, as well as AI-based color correction. Lightroom is getting AI storyboarding and animated text, while in Illustrator, there’s a new generative recolor function that can change the colors of existing images. The company’s all-in-one free tool, Adobe Express, is getting your basic text-to-image generative AI. There is also an AI chatbot interface “co-pilot” that uses text prompts to modify images, without the need for users to take direct control.
Firefly is being integrated into Adobe Express and Creative Cloud apps. The web-based Firefly app is also coming out of beta. More importantly, every customer is being offered a number of “Generative Credits” as part of their plans. Firefly Web App, Express Premium, and Creative Cloud paid subscriptions offer multiple credits for creating content based on text-based prompts.
Most online AI art generators, especially those using “free” services, use a credit system to limit the number of images each user can create. Adobe’s Creative Cloud already costs $US55 a month if you want access to all the company’s apps, while a Photoshop subscription will run users $US21 a month. According to Adobe’s billing page, Photoshop users get 500 credits a month, while Creative Cloud subscribers get 1,000. Users can also pay $US5 per month just for Adobe Firefly, which gives them just 100 monthly credits.
One of the biggest considerations of an AI image generator is the training data used to help create the content. Adobe still says that, unlike OpenAI’s major hitters like DALL-E 2, Midjourney or Stability XL, its text-to-image AI is only using images from its own Adobe stock service. As for compensation from all the creators it openly borrows from, Adobe said that all Adobe stock contributors with content in the Standard or Premium collections will receive a bonus once a year that depends on the total number of approved images and assets. The company did not disclose the value of this bonus. We’ve reached out to Adobe for clarification, and we’ll update this piece if we hear more.
The company claims it has its own system to label what work was created or modified by AI. This so-called “nutrition label” is automatically applied to the metadata of AI content when created by Adobe’s software suite. The Material Authenticity Initiative label includes the property’s name, production data, and the tools used to produce it. It’s powered by an open-source tool, and Adobe promises that “this data will remain relevant to the content wherever it is used, published or stored.” There is also a verification website where users can upload files to see their AI metadata.
This is very different from AI image watermarking, which processes an imperceptible imprint on the pixels of an image. Google claims it has a watermarking process that cannot be modified by filters or other image enhancements but has yet to explain how it works. Instead, Adobe’s label is a simple metadata tag.
The Content Authenticity Initiative is an industry association founded by Adobe, The New York Times, and Twitter. Although metadata is attached to an image after it is created, it is very easy to remove. CAI acknowledged this in its FAQ page, though the association said users can search verified sites to find possible matches with the allegedly modified image.