Introduction

Adobe is doubling down on artificial intelligence across its creative software suite. At its MAX conference in Miami, the company introduced new generative capabilities for Photoshop, Premiere, Illustrator, InDesign, and Firefly, while also opening its ecosystem to outside AI assistants through a new technical standard.

What Happened

The announcements span desktop and web applications. In Photoshop, users will find a relighting feature and adjustable sliders for prompt-based generative styles. Premiere is receiving a professional color grading system and a video object removal tool. Lightroom gains automatic eye-blink correction and blemish removal.

Firefly, Adobe's AI creative platform, is expanding in several directions. A new private beta feature called Builder lets users mock up designs and prototypes, positioning it against tools like Figma. The platform can now generate on-brand charts, edit specific regions of an image while leaving surrounding elements untouched, and assemble short videos from clips, images, and sound effects.

Adobe is also making its creative agent more proactive, allowing it to stay active and suggest quick actions such as background removal or video thumbnail generation. Third-party AI models from Google, OpenAI, Runway, and others remain available, and recent acquisition Topaz Labs is contributing Wonder 4 for upscaling and denoising, plus ImageHDR for converting standard photos to HDR.

Perhaps the most significant infrastructure update is the release of a local Model Context Protocol (MCP). This allows external chatbots including ChatGPT, Claude, and Gemini to access functions inside Illustrator, Photoshop, After Effects, and Premiere. Adobe Vice President Deepa Subramaniam said the company is still determining which capabilities to expose and plans to shape the rollout based on user feedback.

Why This Matters

The move signals Adobe's intent to keep its creative tools central as AI assistants become more capable. By offering an MCP, Adobe is letting outside AI platforms reach into its applications rather than keeping everything inside its own interface. That could change how designers and video editors work, shifting some tasks from manual clicks to conversational commands.

The timing is notable because Figma co-founder Dylan Field recently questioned whether MCP is a mature standard, suggesting it primarily benefits leading AI labs. Adobe's decision to proceed anyway shows confidence that creators will want direct, programmable access to Creative Cloud features.

Key Takeaways

  • Adobe announced new AI features at MAX in Miami for Photoshop, Premiere, Illustrator, InDesign, Firefly, Express, and Lightroom.
  • Photoshop gets relighting controls and prompt-based style sliders; Premiere gains professional color grading and video object removal.
  • Firefly introduces Builder, a private beta prototyping tool, plus region-specific image editing and AI-assisted video assembly.
  • A new local Model Context Protocol lets external chatbots such as ChatGPT, Claude, and Gemini interact with Creative Cloud apps.
  • Adobe says it will refine MCP capabilities over time based on how customers actually use them.

Conclusion

Adobe's latest push blends familiar creative upgrades with a broader bet on AI interoperability. The new editing tools should help everyday users produce polished images and video faster, while the MCP could reshape how professional workflows connect to outside AI assistants. As with any early-stage protocol, the real test will be whether developers and creators adopt it in large numbers.