Adobe Premiere Pro to feature new generative AI-powered video editing tools

Adobe to launch a new tool called content credentials to identify whether the material is genuine or created by AI
An undated image showing a man using Adobe Premiere Pro. — Unsplash
An undated image showing a man using Adobe Premiere Pro. — Unsplash

Adobe, the software giant unveiled new tools for the popular video editing software Premiere Pro, powered by generative AI technology.

Editors can easily remove or add an item to a movie by providing a simple text prompt powered by Firefly AI, which includes a new video creation model. These new tools will be accessible to Premiere Pro customers starting in May.

Moreover, Adobe is collaborating with OpenAI and Runway to include more AI-backed video editing tools in Premiere Pro, which includes the capacity to create a B-roll and the generative extend tool, which allows editors to add a few seconds to film.

Premiere Pro's new object addition or removal tool allows editors to quickly select, track, and remove unwanted objects from the frame with a single text prompt, such as "change an actor's wardrobe" or "quickly add set dressings such as a painting or photo-realistic flowers on a desk."

Moreover, Adobe will launch a new tool called content credentials, which uses a watermark to easily identify whether the material is genuine or created by AI.

The new generative extend function may add frames to a movie to make it longer, which is useful for tasks like fine-tuning edits and creating a better transition between material. This is driven by both in-house Firefly models and third-party models such as Pika, OpenAI's Sora, and the Runway big language model.

The company demonstrated a new text-to-video model that can produce whole new material within Premiere Pro by simply typing a text prompt or supplying a reference picture.

The software giant also demonstrated new audio operations, such as interactive fade handles, essential sound badges with audio categories, effect badges, and a revamped waveform in the timeline of the app.