Adobe has introduced a new AI Assistant in Photoshop that lets users make image edits just by describing them in everyday language. Announced at the company’s annual Adobe MAX 2025 event on Tuesday, the new conversational tool is powered by Adobe’s Firefly Image Model 5. According to the company, using the AI Assistant, instead of relying on complex tools or layers, users will now be able to simply type prompts like remove the background,” “make the lighting softer,” or “add a reflection,” and the AI will carry out those edits while keeping the rest of the image intact.
Adobe notes that this new feature called Prompt to Edit, works like a text-to-image generator. However, instead of creating a new picture, it modifies existing ones based on written prompts. Giving an example of how it can be used, Adobe explains that users can type “brighten the sky and make it sunset orange,” and Photoshop will generate the edit while preserving the rest of the image.
This new AI tool is part of the company’s growing push towards agentic AI, aimed at simplifying creation for both professionals and casual users.
During the event, the company also previewed its Project Moonlight, a conversational assistant that can work across Photoshop, Firefly, and other Adobe apps to offer creative suggestions, content insights, and workflow support.
Notably, the new Photoshop AI capabilities are built on Adobe’s Firefly Image Model 5, which is now available in public beta. According to the company, it is designed to handle complex, multi-layered compositions with greater photorealism. It can generate images in native 4MP resolution and maintain consistency in lighting, texture, and movement.
Adobe also revealed that more tools are on the way. Upcoming features like Layered Image Editing, currently in development, will further improve precision by keeping every change context-aware.
New AI features in Firefly
Alongside the Photoshop upgrade, Adobe also announced a major update to its Firefly creative AI studio. The studio now features expanded AI capabilities, bringing together image, video, and audio generation tools in one place. The platform’s latest additions include Generate Soundtrack for creating fully licensed, AI-generated music, and Generate Speech for producing realistic voiceovers in multiple languages. Both tools are in public beta and can be synced directly with video projects. Adobe also unveiled a timeline-based AI video editor, now in private beta, that combines traditional editing with generative features, allowing users to trim, sequence, and edit clips simply through text instructions.
Firefly’s new Custom Models, also available through a private beta, will let creators train their own AI models on personal styles or brand guidelines to produce consistent visual outputs. Adobe announced that the ecosystem is expanding to include partner models from Google, OpenAI, ElevenLabs, and Topaz Labs, giving users access to a wide range of creative AI systems within Adobe’s environment. In addition, Firefly Boards which is the platform’s AI-powered ideation tool has been updated with new 3D rotation, preset generation, and bulk export features to make brainstorming and collaboration easier.
Availability of new AI features by Adobe
Most of these features are rolling out in stages. The Prompt to Edit feature and Firefly Image Model 5 are already live, while the new video and custom model tools will begin rolling out to early-access users next month.
– Ends
Source link
[ad_3]
