At its Max conference, Adobe unveiled a series of experimental AI-powered tools, dubbed “sneaks,” that showcase new, intuitive ways to edit photos, videos, and audio. These prototypes highlight the company’s push to make creative workflows faster and more accessible through generative AI. One of the most striking demonstrations was Project Frame Forward, which enables video editors to make changes to an entire video based on just a single frame. Instead of manually applying masks to select people or objects—a tedious process—Frame Forward automatically detects and removes or replaces elements across the full clip. In a demo, the tool identified and erased a woman from the first frame and seamlessly replaced her with a background that matched the scene’s context, similar to Photoshop’s Context-Aware Fill. Users can also insert new objects by drawing where they want them and describing what to add using AI prompts. The system intelligently applies these changes throughout the video, even ensuring contextual accuracy—such as generating a puddle that reflects the movement of a cat already in the scene. Another standout is Project Light Touch, which uses generative AI to manipulate lighting in images with unprecedented control. Editors can shift the direction of light, simulate lamps that weren’t originally on, or adjust how light diffuses across surfaces. The tool allows users to drag light sources across the canvas in real time, bending light around people and objects. For instance, it can illuminate a pumpkin from within or transform a daytime scene into night, complete with dynamic shadows and color shifts. The warmth or tone of the light can be fine-tuned, enabling effects like soft golden glows or vibrant RGB lighting. Project Clean Take focuses on audio editing, using AI to correct or rephrase spoken words without requiring a re-recording. Users can modify how a speaker enunciates, change the tone to sound more cheerful or curious, or replace specific words while preserving the speaker’s unique vocal characteristics. The tool also separates background sounds into individual sources, allowing editors to mute or adjust specific noises—like a fan or traffic—while keeping the original voice clear and natural. Other experimental tools on display include Project Surface Swap, which lets users instantly change the material or texture of surfaces, and Project Turn Style, which allows 3D-like rotation of objects in 2D images. Project New Depths enables depth-aware editing in photos, ensuring that inserted objects appear partially hidden behind other elements, as if they exist in a true 3D space. While these tools are not yet available to the public and are not guaranteed to become official features, many have roots in past sneaks that later evolved into core tools—such as Photoshop’s Distraction Removal and Harmonize. Adobe’s sneak previews offer a glimpse into the future of creative software, where AI helps artists achieve complex edits with minimal effort. More details can be found on Adobe’s official blog.