AI took centre stage at Adobe MAX LA 2025

When Adobe MAX LA 2025 opened its doors at the end of October, alongside audiences tuning in from around the world, creatives were eager to see exactly where Adobe would take their AI features. It was clear from the get-go that Adobe is hoping designers, artists and illustrators recognise that its integration of AI across their ecosystem is an incredibly powerful tool, rather than a creative replacement – it listens to plain language instructions and quietly chips away at the repetitive aspects of making work. 

The most notable inclusion is the integration of conversational agents within Photoshop, Firefly, and Express. Rather than hunting through menus, you can describe what you’re trying to achieve – be it adjusting lighting, preparing a batch automation, cleaning up an image – and the assistant will line up the relevant tools, effects or models to get you close, ready for you to refine by hand. In Express, that same idea is tuned for people who might not identify as designers at all: the new AI Assistant, currently in beta, can restyle a template, whilst also tweaking copy, swapping out imagery and adjusting specific layers based on loose prompts.

On the note of image creation, Firefly, Adobe’s AI environment, has also levelled up. The latest Firefly Image Model 5 is designed to generate more detailed, higher-resolution imagery and to handle edits described in everyday language. This means you can nudge an existing image rather than starting over each time. Around it sits an increasingly busy toolkit, including Firefly Boards for collaborative moodboarding and end-to-end audio and video features, such as Generate Soundtrack and Generate Speech. 

Inside Creative Cloud, the updates are more targeted, such as Photoshop’s refined Generative Fill, which is powered by a combination of Firefly and partner models, including Generative Upscale and Harmonize. This means one can use low-resolution or composited images with a quality closer to the final result. Premiere Pro equally got an AI upgrade, with AI Object Mask. Currently in beta, AI Object Mask tracks people or objects across a shot, meaning you can grade or blur them without manual rotoscoping. None of these tools changes what a finished image or film needs to be, but they do squeeze some of the time out of getting there. 

For studios and brands working at scale, Adobe also used MAX to talk about infrastructure. GenStudio, its content platform, now threads generative features and AI agents more directly into the content supply chain – from planning and asset creation through to publishing. Meanwhile, Firefly Foundry allows companies to work with Adobe on custom models trained on their own IP, so future campaigns can be built from an engine that is aware of the brand’s visual language. 

Taking a step back and looking at the three-day event, the announcements in LA feel like important practical progress, whereby the decisions, practice and process behind creativity belong squarely to the person behind the keyboard. Adobe’s approach to automation hopes to simply free up the time for the parts of the process that can’t be automated. Head to Adobe MAX Online to rewatch keynotes and Sneaks, catch any sessions you missed, and explore the new features in Firefly, Creative Cloud and Adobe Express as they continue to roll out.