Adobe is bringing a new wave of generative AI tools directly into Premiere and After Effects, giving video editors more ways to generate content, clean up audio and automate repetitive tasks without leaving their editing environment.

The updates, announced on September 8 ahead of IBC 2026, are centered around keeping editors in the timeline. Premiere is gaining a new Generative Media Tool for creating video and audio, while After Effects is expanding its AI Assistant into public beta for motion-design workflows.

For Adobe, the goal isn’t simply to automate video production. The company is positioning these tools as assistants that handle tedious work while editors retain control over the final result.

Premiere gets a Generative Media Tool

The biggest addition to Premiere is the new Generative Media Tool, which allows editors to generate video and sound effects directly within the timeline.

Instead of stopping an edit to search through stock libraries or another generative AI service, an editor can select where new content should appear, describe what is needed and generate an editable clip in context. Premiere can also use reference frames from the existing project when generating video, helping the resulting footage better match the surrounding edit.

Adobe is giving editors a choice of models, too. The tool supports Adobe Firefly as well as partner models from Google Veo, Kling, Runway and Luma.

The Generative Media Tool includes several individual capabilities. Generate Video can fill gaps in a timeline, while Generate Sound Effects creates effects directly in the project. Editors can even use their voice to guide the rhythm, timing and intensity of generated sound effects.

Premiere is also getting Generate Soundscape in beta. The feature can analyze up to 15 seconds of video and create a contextual audio layer containing ambience and sound effects timed to the edit.

Generate Music, also in beta, can create original music with multiple variations and controls for tempo and looping. Adobe says the generated music is commercially safe and includes synchronization rights across different types of media.

That combination could make generative AI considerably more practical for everyday editing. Rather than generating an entire finished video, editors can use AI for the small missing pieces that can otherwise interrupt a workflow.

Premiere’s audio tools get smarter

Adobe is also adding a collection of AI-powered audio features to Premiere in public beta.

Enhance Audio builds on Adobe’s existing Enhance Speech technology with more advanced noise reduction and independent controls for dialogue, music, ambience and sound effects. Editors can adjust those elements separately within a clip or remove individual components entirely.

Separate Crosstalk addresses another common problem: two people speaking over one another. The feature can automatically separate overlapping speakers into independent tracks, allowing editors to mix and edit them individually without having to record the conversation again.

Dynamic Auto Ducking can automatically balance dialogue against music and other audio elements, reducing the need to create manual keyframes.

Adobe is also introducing a new speed algorithm designed to make duration changes sound more natural across a wider range while preserving the original pitch and timbre.

Together, these tools target some of the less glamorous parts of post-production—the cleanup and adjustment work that can consume a surprising amount of editing time.

After Effects adds AI Assistant in public beta

Adobe is extending its AI Assistant beyond Premiere, with the feature now available in public beta in After Effects.

The assistant is designed to handle multi-step and technical tasks through natural-language instructions. Adobe says it can work across an entire After Effects project rather than being limited to the composition currently open.

That means editors and motion designers can ask the assistant to reorganize large projects containing thousands of layers, troubleshoot technical problems, inspect compositions and footage, and create motion effects such as expressions without having to write the code themselves.

The idea is less about handing an entire project over to AI and more about delegating the repetitive work. Adobe emphasizes that users decide what the assistant should handle and remain responsible for refining and approving the resulting editable work.

For experienced After Effects users, that could make the feature particularly useful for project maintenance and technical tasks that are necessary but don’t necessarily require hands-on creative work.

For newer users, meanwhile, natural-language instructions could provide a more approachable way into some of After Effects’ notoriously deep capabilities.

Premiere also gets new color tools

Adobe is continuing to develop Color Mode in Premiere, which remains in beta.

The editor-focused grading experience is gaining new Style Modules, including Vignette and Midtones, along with an Auto Color tool that can correct color temperature and exposure with a single click.

Auto Color can be applied to an individual clip, a custom group or an entire sequence. Adobe says the broader goal is to give editors more professional color-control tools without requiring them to move into a separate grading workflow.

Frame.io storage comes directly to the timeline

There are some less flashy workflow improvements in the update as well.

Adobe has redesigned Premiere’s Properties and Effects panels to make navigation faster. Frame.io Mounted Storage is also now available to all accounts, allowing users to stream full-resolution media from the cloud directly into a Premiere timeline without manually downloading files first.

For projects involving large amounts of footage, that could help reduce the need for local storage and eliminate another interruption between finding media and actually using it.

Adobe’s latest Premiere and After Effects updates show where its video strategy is heading: generative AI is becoming less of a separate destination and more of a native layer inside established creative applications.

The most interesting part may be that Adobe isn’t limiting AI to flashy generation features. Filling a missing shot is useful, but so is separating overlapping dialogue, balancing audio, reorganizing thousands of layers or fixing a technical problem through a plain-language command.

That approach could make AI feel less like a novelty and more like another set of tools sitting alongside the traditional editing controls—exactly where professional editors already spend their time.

The Premiere announcement also comes as Adobe celebrates the software receiving an Engineering, Science & Technology Emmy Award at the 2026 Emmys, recognizing 35 years of Premiere’s role in television production.