Adobe brings AI agents and new creative tools to Photoshop, Premiere, and Firefly
AI-assisted digest, checked against the original source before publishing. How we work
Adobe's MAX announcements push AI agents, prototyping, and a local MCP server across its creative suite.
Adobe used its MAX event in Miami to roll out a wide set of AI additions across Photoshop, Premiere, Illustrator, and InDesign, while expanding Firefly and releasing a Model Context Protocol implementation that lets external chatbots such as ChatGPT, Claude, and Gemini invoke Creative Cloud features.
Firefly's most competitive move is Builder, a prototyping tool in private beta that pushes Adobe into Figma's territory. The app also gains a tool for producing on-brand graphs and an image editor that alters one part of a picture while leaving other objects intact. A new AI video tool can assemble a finished piece from clips plus elements like images and sound effects. Adobe is also making its agent proactive and always on so it can gather more context about a user's work and be present for any action, complemented by quick actions such as background removal or generating a video thumbnail.
In the desktop apps, Photoshop gains a relighting feature and sliders for prompt-based generative styles; Premiere adds a professional color grading system and an object removal tool for video; and Lightroom gains blink correction and a blemish removal tool. Adobe framed the MAX announcements as a continuation of AI features it has already layered into Photoshop, Express, Premiere, and Illustrator.
On the model layer, Adobe continues to draw on third-party providers including Google, Runway, ElevenLabs, Black Forest Labs, OpenAI, Kling, and ByteDance. Its newer Topaz Labs acquisition is contributing two image models: Wonder 4, aimed at upscaling and denoising, and ImageHDR, which converts standard images into HDR.
The MCP server is local and widens what outside chatbots can reach, building on existing ChatGPT and Claude integrations. Adobe vice president Deepa Subramaniam said the company is still deciding which endpoints to expose and will let user behavior shape that decision, describing an approach of publishing a number of endpoints across Illustrator, Photoshop, After Effects, and Premiere, then watching what people attempt before revealing more, with no current intent to constrain what the endpoint can expose.
What to watch: whether Builder graduates from private beta and how it lands against Figma, which co-founder Dylan Field argued last month is not well served by MCP as a standard that benefits frontier labs more. Open questions include how Adobe will scope endpoints as usage emerges, and what operational tradeoffs come with an always-on agent that stays present across a creative workflow.
Key facts
- • Adobe announced the AI features at its MAX event in Miami.
- • Firefly's Builder prototyping tool is in private beta.
- • A local MCP exposes Creative Cloud features to ChatGPT, Claude, Gemini.
- • Adobe's Topaz Labs acquisition ships Wonder 4 and ImageHDR models.
- • Adobe VP Deepa Subramaniam said endpoint scope will follow user feedback.
Why it matters
Builders get a sanctioned programmatic surface into Photoshop, Premiere, Illustrator, and After Effects, so creative actions can be scripted from external agents rather than clicked. Adobe's open-ended MCP stance, contrasted with Figma's public skepticism of the standard, suggests a land-grab over which tool owns the agent-facing creative workflow. Watch which endpoints ship next and whether the a
Spotted an error? Send a correction.
Get the day's AI news in one email
The Daily Epoch · 60-second read · unsubscribe anytime
