Adobe Firefly – ethical text-to-image AI announced

Adobe Firefly - ethical text-to-image AI announced


Adobe’s Creative Cloud applications already use artificial intelligence but the company has announced the introduction of a new text-to-image generator called Firefly that will be part of its Sensei technology.

According to Adobe, Firefly will bring precision, power and speed a to Creative Cloud, Document Cloud, Experience Cloud and Adobe Express workflows where content is created and edited. One of the biggest concerns surrounding the use of AI in content creation is the source of the images that are used. Adobe aims to protect creators while still empowering them with the latest technology.

Consequently, Firefly’s first training model will use Adobe Stock images, open-licensed content and public domain content outside of copyright. The aim is to enable the creation of content that is suitable for commercial use. Adobe is also set to introduce a ‘Do not train’ tag that can be added to images that creators want excluded from any AI training. The tag will stay with the content no matter where it is used.

Adobe also plans to allow creators to train Firefly with their own content, thus enabling users to create new content in their own style.

A public beta program has been announced for Firefly and creators with an Adobe account are invited to request to get involved with access to the features that are in development. There will also the opportunity to ask the development team questions and give feedback.

Follow the link to request an Adobe Firefly invite.