At Adobe Max 2026 in Miami today, we got glimpses of where Adobe's creative software could go next with the traditional Adobe Sneaks reveals. Always a highlight of the annual event, the teasers reveal the tech and tools that could come to Adobe apps sometime in the future (and they do tend to arrive in some form).
Unsurprisingly, this year's Sneaks have a heavy focus on AI across the ten innovations. We saw demos of tools that could make AI image and video editing more precise and transform the workflows of people working in film, digital art, photography and design. Here are the highlights.
Project Scene Sweep
On the video and filmmaking front, Adobe's revealed what it's calling Project Scene Sweep. The AI-driven tool is being designed to understand characters, locations and scenes across a sequence, helping creators apply an edit consistently across cuts and camera angles.
It allows long-form, multi-shot videos to be edited from a single text prompt or edited frame, with the changes applied across each shot, cut, and camera angle. Example use cases cited include post-production fixes like removing a person or prop or swapping wardrobe without a reshoot.
The tool could also help for clearance issues, for example to replace an unlicensed logo or artwork in every shot where it appears, or for product placement after a shoot. Adobe suggests it could even be used to recast a new lead or location for a video or for testing variants for localisation.
Other projects in the works for video editing are Project Sound Bender, which would allow users to describe a sound and pull it out of a finished mix, then explore variations on an interactive canvas, and Project Live Set, which would allow filmmakers to change a subject or environment or redirect the camera as a scene continues.
Project Deep Brush
For digital art and image editing, Project Deep Brush uses AI to understand an image’s shape, colour and lighting and separate images into physically decomposed layers. This would give artists a new level of control for painting and retouching, including the ability to control where paint lands and to make lighting-aware edits in photos.
Users would be able to add things like graffiti, signage, wraps, or branding to surfaces like walls, pillars, and vehicles, making the tool a potential game changer for product and environment mockups. For example, you could put a logo or texture on a curved or angled surface like a bottle, and the edit would follow the surface’s contours and lighting.
You'll be able to try this out for yourself on the new Adobe Labs Playground site.
Project Hands On
Project Hands On is presented as a possible step from text-based editing towards a more fully graphical interface allowing greater control of AI-generated imagery. It analyses what’s in the image and then adds controls, such as control points that can be used to reshape details by simply dragging rather than rewriting a text prompt.
This could make editing AI-generated output more precise and intuitive, avoiding the trial-and-error nature of having to repeatedly rewrite prompts and regenerate an image.
Project Screen Test
Project Screen Test was perhaps the wildest idea teased at Adobe Sneaks. It would allow character designers to talk with their fictional creations to learn more about them and how they might behave in a given situation. Yes, seriously!
The tool would allow creators to develop a character’s look, voice and personality, and then have a natural language dialogue with said character in order to explore who they are and how they might respond in different scenes. The character would effectively audition for you.
The idea is that this conversation could turn character development in generative filmmaking and game development into a more intuitive, expressive part of the storytelling process.
Project Sight Span
Project Sight Span is tool that feels a lot more familiar. It transforms an image into an interactive, navigable world that creators can step inside and explore freely. That exploration could be used to frame shots beyond the original image, and render new images or camera-controlled videos, a little like what some artists are already doing with Gaussian splats. Adobe reckons this could be useful for things like set design and world building.
Project Link Up
Another highlight of Adobe Sneaks 2026 was Project Link Up. With existing vector tools, dragging an object onto another only determines where it lands. Project Link Up turns that drag gesture into semantic fusion, combining an understanding of both objects with drag path, velocity, and creative controls to shape how they influence each other in real time, while keeping the result fully editable as vector artwork.
The idea is that artists and designers would be able to place 2D or vector assets into scenes with automatic perspective matching, reducing the need to redraw objects at new viewing angles. For typography, it would be possible to merge illustrated elements with letterforms while keeping the type editable for branding, logotypes, and packaging design.
For more news from Adobe MAX, see what's new in Photoshop 2027.
