Adobe Launches Photoshop AI Assistant in Beta and Expands Firefly Media Suite with Advanced Generative Editing

Adobe debuts a Photoshop AI assistant for web and mobile, enabling natural language editing, AI markup tools, and expanded generative features in Firefly.

By: AXL Media

Published: Mar 10, 2026, 9:13 AM EDT

Source: The information in this article was sourced from TechCrunch

Generative AI Assistant Enters Photoshop Ecosystem

Adobe has moved its highly anticipated AI assistant for Photoshop into public beta, extending the feature to both web and mobile applications. Initially previewed at the Adobe MAX event in October, this assistant serves as a natural language interface that allows creators to execute sophisticated edits without manual tool selection. Users can now interact with the software using plain English to perform tasks such as removing specific people from a frame, altering the color temperature of a scene, or adjusting lighting conditions. This shift marks a significant transition from traditional menu based editing toward a conversational, intent-driven creative process.

Natural Language Prompts for Technical Execution

The AI assistant is designed to handle a wide array of technical adjustments that previously required significant manual expertise. By providing simple instructions, users can direct the AI to add a soft glow to a subject, crop an image into specific social media formats, or enhance existing shadows for better depth. Additionally, the assistant can transform entire backgrounds based on descriptive prompts, allowing for rapid prototyping of different visual looks. This capability is intended to lower the barrier to entry for novice editors while providing a high speed shortcut for professional designers working on iterative projects.

Introduction of AI Markup and Interaction Tools

Adobe is further innovating the user interface with a new "AI markup" feature currently in public beta. This tool enables users to draw physical markers or sketches directly onto the screen, which the AI then interprets as instructions for transformation. For example, a user can sketch a rough shape of a flower or circle a specific object to signify it should be modified or removed. This hybrid approach combines traditional stylus or touch input with generative intelligence, providing more granular control over exactly where and how the AI applies its edits within a composition.