Adobe Expands AI Agents Across Creative Cloud, Enters Public Beta

Adobe has launched a public beta of its expanded AI agents across all major Creative Cloud applications. The assistant, built on what the company calls a creative agent, is now available in Photoshop, Premiere Pro, Illustrator, InDesign and Frame.io, while After Effects will get it in a closed release for now. It's the largest AI integration yet into the suite of tools used by millions of designers, editors and creators worldwide.

What the Agent Does in Practice

The capabilities vary by application. In Photoshop, the agent manages layers, batch-removes backgrounds and swaps out photo backdrops. In Premiere Pro, it sorts footage into libraries, automatically renames files, recognizes interview questions and adds markers to speed up editing. Illustrator gained multi-step file generation, layer reorganization and automatic pre-press checks, while InDesign can update templates across multiple layouts at once while keeping styles and print settings consistent.

Frame.io, Adobe's collaboration platform for video assets, gained features for organizing on-set footage, capturing team feedback and generating supplemental shots. Firefly, Adobe's generative platform, received expanded tools for building brand identities, premium-style short product videos, automatic dialogue-driven editing, and storyboards that convert directly into video.

The Happy Path Philosophy

Deepa Subramaniam, Adobe's vice president of creative product marketing, described the company's approach as guiding users down a "happy path" rather than leaving them alone with a blank prompt field. The agent is meant to orchestrate multi-step tasks, but it's the creator who decides what to hand off to the machine, what to keep for themselves, and how to use their own taste and experience to shape the final result.

Every creator now has an agent that can help them work across every app and every platform they use. - David Wadhwani, President of Creativity & Productivity at Adobe

Market Context

Adobe first announced these features in April 2026, and June's public beta launch delivers on those plans at a broader scale. The company has spent years adding individual machine-learning-based tools to its applications, but has rarely allowed AI to independently carry out multi-step production tasks without constant user oversight at every step.

The move fits into a broader trend of agentification across creative and office tools, as software makers shift from single generative prompts to assistants that execute entire chains of tasks. Integration with external chat platforms, including ChatGPT, Claude, Copilot and Gemini, is meant to let users trigger Creative Cloud tasks directly from those tools without opening Adobe's own applications.

What This Means for Users

For freelancers and design studios, this could mean less time spent on repetitive tasks such as organizing files, batch-removing backgrounds or sorting video footage. Adobe stresses, though, that its internal research shows clear resistance among creators to handing the machine the final say on artistic matters, which explains the decision to keep the agent in an executor role rather than a decision-maker one.

The public beta means the features are already available to Creative Cloud subscribers, though they may still change before full rollout. After Effects and the updated Firefly studio remain closed for now, accessible only via waitlist signup, suggesting Adobe is testing these features more cautiously than in the apps already made available.