Adobe is in the process of developing an AI model designed to generate video content, although the company has not disclosed a specific launch date for this project.
Positioned as a response to offerings such as OpenAI’s Sora, Google’s Imagen 2, and similar models emerging from startups in the burgeoning generative AI video sector, Adobe’s initiative is part of its expanding Firefly suite of generative AI products.
Expected to integrate into Premiere Pro, Adobe’s flagship video editing software, later this year, the AI model will introduce three new features: object addition, object removal, and generative extend.
Like many other generative AI video tools available today, Adobe’s model has the capability to create footage from scratch, utilizing prompts or reference images as input.
In response to concerns regarding the potential for deepfakes associated with generative AI tools, Adobe is implementing Content Credentials in Premiere Pro.
These metadata tags, aligned with Adobe’s Content Authenticity Initiative, already present in Photoshop and part of Adobe’s image-generating Firefly models, will not only identify AI-generated content but also specify the AI model responsible for its creation.
Regarding pricing, Adobe has not disclosed the cost for customers to access the upcoming video generation features in Premiere Pro, indicating that pricing details are still under consideration.
However, the payment structure will follow the generative credits system established with Adobe’s earlier Firefly models. Under this system, customers with a paid subscription to Adobe Creative Cloud receive a monthly allotment of generative credits, ranging from 25 to 1,000 credits per month depending on the subscription plan.
These credits renew at the beginning of each month, enabling users to access the generative AI capabilities within Adobe’s suite of products.
