Adobe unveiled transformative AI capabilities at the Adobe MAX 2025 conference on Tuesday, introducing conversational assistants across its flagship creative applications while partnering with Google Cloud and OpenAI to integrate industry-leading AI models directly into tools like Photoshop and Premiere. The updates represent a fundamental shift in how creators interact with professional editing software, moving from manual controls to natural language conversations.
Table of Contents
What Are Adobe’s New AI Assistants?
Adobe’s new AI assistants enable users to control creative applications through conversational commands, eliminating the need to navigate complex menus and toolbars. The assistants can perform multi-step editing tasks by manipulating buttons, sliders, and panels automatically while users describe their desired outcomes in plain language.
Prompt-to-Edit Feature in Photoshop
The centerpiece of Photoshop’s AI enhancement is the “Prompt to Edit” feature, powered by the new Firefly Image Model 5. Users can type natural language commands like “brighten the sky and make it sunset orange” or “remove the background,” and the AI assistant executes these complex edits automatically. When switching to Photoshop’s “agentic” mode, the traditional interface minimizes to display a simple prompt-based UI, allowing creators to focus on creative direction rather than technical execution.
AI Agents Across Express and Firefly
Adobe Express and the Firefly platform now feature conversational AI agents that can generate images, change backgrounds, modify text, and provide personalized recommendations. The Express AI assistant helps users easily navigate the app’s capabilities while offering tutorials for complex tasks. Unlike rigid automation tools, these agents allow users to seamlessly switch between conversational prompts and manual fine-tuning using traditional sliders and adjustment tools.
Google Cloud Partnership
Adobe’s expanded collaboration with Google Cloud brings three of Google’s most advanced AI models Gemini, Veo, and Imagen directly into Adobe’s creative ecosystem. This integration provides creators with choice and flexibility, allowing them to select the best AI model for specific tasks without leaving Adobe’s secure platform.
Gemini, Veo, and Imagen Integration
The partnership integrates Google’s models across Adobe Firefly, Photoshop, Adobe Express, Premiere, and GenStudio. Veo specializes in video generation, Imagen focuses on photorealistic image creation, and Gemini powers multimodal AI capabilities. Users can access these models alongside Adobe’s proprietary Firefly models, creating richer images and higher-quality video content through a unified interface.
Enterprise Customization Through Firefly Foundry
Enterprise customers gain the ability to build brand-specific AI models through Adobe Firefly Foundry using proprietary data. These custom models leverage Google Cloud’s Vertex AI platform, ensuring data security and privacy while enabling personalized content creation at scale. Adobe’s Chairman and CEO Shantanu Narayen stated that the partnership “brings together Adobe’s creative DNA and Google’s AI models to empower creators and brands to push the boundaries of what’s possible”.
Firefly Image Model 5 Capabilities
Adobe introduced Firefly Image Model 5 as its most advanced AI imaging model, combining photorealistic quality with groundbreaking editing capabilities. The model represents a significant leap forward in resolution, realism, and creative control compared to previous generations.
Native 4-Megapixel Resolution
Firefly Image Model 5 generates images at native 4-megapixel resolution, eliminating the upscaling requirements that plagued earlier models. The enhanced resolution delivers stronger photorealism, improved human anatomy rendering, and better handling of lighting, texture, and environmental details. This native high-resolution output puts Firefly on competitive footing with other leading AI image generators in terms of technical quality.
Layered Editing Technology
The model’s innovative layered editing capability treats image elements as separate layers, enabling users to resize, rotate, or modify individual objects while the AI automatically adjusts shadows, lighting, and perspective. The Layered Image Editing feature, currently in development, brings content-aware adjustments that maintain visual coherence as creators manipulate different elements. This approach combines the flexibility of traditional layer-based editing with AI-powered automation.
Audio and Video AI Tools
Adobe expanded Firefly’s capabilities beyond imaging to include studio-quality audio and video generation tools designed for professional content creators.
Generate Soundtrack
The Generate Soundtrack feature creates fully licensed audio tracks suitable for commercial projects. This tool addresses a persistent challenge for video creators who need royalty-free music that matches their content’s mood and pacing. The commercially safe licensing model ensures creators can use generated soundtracks without copyright concerns.
Generate Speech for Multilingual Content
Generate Speech provides multilingual voiceover capabilities, allowing creators to produce professional-quality narration in multiple languages. This feature streamlines localization workflows for global content distribution and makes professional audio production accessible to creators without recording equipment or voice talent.
ChatGPT and YouTube Integrations
Adobe announced strategic partnerships extending its creative tools into new platforms where creators already spend time.
OpenAI Partnership Details
Adobe is developing integration with OpenAI’s ChatGPT that allows users to start creative projects directly within the chatbot interface. Users can initiate a conversation in ChatGPT to conceptualize and begin designing content, then transition seamlessly to Adobe Express for advanced editing using the full creative toolkit. Ely Greenfield, Adobe’s Chief Technology Officer for Digital Media, explained that users can “modify the adjustments the agent is making” and “transfer into the app” to access “the complete range of editing tools”. While the AI assistant in Express is currently accessible, Adobe has not specified a release date for the ChatGPT integration.
Create for YouTube Shorts Feature
Adobe partnered with YouTube to bring professional video editing capabilities directly into YouTube Shorts through a new “Create for YouTube Shorts” feature in the Premiere mobile app. The integration creates a dedicated content creation space where creators can produce and instantly publish videos to YouTube Shorts without leaving the Premiere environment. Users access this functionality by tapping an “Edit in Adobe Premiere” icon within YouTube.
What This Means for Creators
These announcements signal Adobe’s strategy to democratize professional creative tools through AI assistance while maintaining the depth required by advanced users. The conversational interface lowers the learning curve for beginners who find traditional creative software intimidating, while layered editing and manual control options preserve the precision that professionals demand.
The multi-model approach integrating Adobe Firefly alongside Google, OpenAI, and other third-party models from Topaz Labs, Runway, and Luma AI provides creators with flexibility to choose optimal tools for specific tasks. This ecosystem strategy contrasts with closed-platform approaches, positioning Adobe as a hub that orchestrates best-in-class AI capabilities.
For enterprise teams, the Firefly Creative Production batch editing solution (currently in private beta) enables processing thousands of images at once with consistent background replacements, color grading, and cropping through a no-code interface. Combined with custom model training through Firefly Foundry, these capabilities support brand-consistent content creation at unprecedented scale.
Comparison Table
| Feature | Adobe Firefly Model 5 | Previous Firefly Models | Third-Party Integration |
|---|---|---|---|
| Native Resolution | 4-megapixel | Lower (required upscaling) | Varies by model |
| Layered Editing | Yes, automatic shadow/light adjustment | No | Limited |
| Conversational Control | Prompt-to-Edit | Basic prompts | Model-dependent |
| Commercial Licensing | Fully licensed | Fully licensed | Varies (check terms) |
| Model Options | Firefly + Google (Gemini, Veo, Imagen) + OpenAI | Firefly only | Through partnerships |
| Enterprise Customization | Firefly Foundry with brand data | Not available | Limited |
Frequently Asked Questions
What is Adobe’s AI assistant and which apps support it?
Adobe’s AI assistant is a conversational interface that lets you control creative applications through natural language commands. It’s currently available in Photoshop, Adobe Express, and Firefly, with the ability to perform complex multi-step tasks like color correction, resizing, and content generation through simple chat interactions.
How does Firefly Image Model 5 differ from previous versions?
Firefly Image Model 5 generates images at native 4-megapixel resolution without upscaling and introduces layered editing capabilities that treat image elements as separate layers. The model provides improved photorealism, better human anatomy rendering, and enhanced lighting and texture quality compared to earlier iterations.
Can I use Google’s AI models in Adobe apps?
Yes, Adobe’s partnership with Google Cloud integrates Gemini, Veo, and Imagen models directly into Firefly, Photoshop, Express, and Premiere. Users can select these models alongside Adobe’s Firefly models for image and video generation tasks.
When will ChatGPT integration with Adobe be available?
Adobe announced the ChatGPT integration at Adobe MAX 2025 but has not provided a specific release date. The integration will allow users to start creative projects in ChatGPT before transitioning to Adobe Express for advanced editing.
Are Adobe’s AI-generated assets commercially licensed?
Yes, Adobe’s Firefly models generate commercially safe content, including images, soundtracks, and speech. This licensing model distinguishes Adobe from some competitors whose AI-generated content may have usage restrictions.
What is Project Moonlight?
Project Moonlight is Adobe’s conversational AI assistant designed to work across creative apps and connect to creators’ social channels. It provides personalized recommendations and multi-step task automation, though Adobe has only previewed this capability without announcing general availability.
Featured Snippet Boxes
What is Adobe’s AI Assistant?
A conversational interface in Photoshop, Express, and Firefly that runs edits from natural language prompts. You can chat with your apps to do color correction, remove objects, and generate images without manually tweaking tools and sliders.
Key Features of Firefly Image Model 5
Generates photorealistic images at native 4 MP with layered editing. It treats elements as separate layers and automatically adjusts shadows and lighting when objects are resized or moved, combining layer-based control with AI automation.
Adobe Google Partnership Benefits
Integrates Google’s Gemini, Veo, and Imagen models into Photoshop, Premiere, Express, and Firefly. Enterprise teams can customize these models with proprietary data via Firefly Foundry, while Vertex AI on Google Cloud provides security and governance.