Adobe MAX 2026: Adobe Goes All In on Agentic AI

I'm spending today and tomorrow at Adobe MAX, a three-day creative conference held in Miami and attended by over 10,000 creatives. Adobe’s goal for the event is to share its latest innovations so that this creative community can best utilize these new tools. One theme apparent at this year’s MAX is removing the barriers caused by complex software interfaces. Adobe’s stated mission is to empower everyone to create, and the widespread implementation of AI across the Creative Cloud suite appears well-suited to that purpose.

I've been using generative AI a lot in my photography over the past few years. This takes the form of adding or removing elements from my photographs with simple commands like "remove the water bottle from the table." I also use generative AI to change elements such as the knot in someone's tie or the type of watch they are wearing. Most of the additions made to the Creative Cloud suite are based on agentic AI, which takes things a step further by allowing software to perform a series of complex tasks from a few simple instructions.

The capabilities of programs like Photoshop have always exceeded our ability to master the tools necessary to make the most of that complexity. Adobe is determined to make human-machine interaction as seamless and smooth as possible through the implementation of a new creative agent in Illustrator, Premiere, InDesign, After Effects, and Firefly. You give clear direction in natural language, and these programs perform multistep tasks.

With agentic AI, you don’t have to know or understand every step necessary to create a desired result. You can start with the end goal and allow the program to create the steps necessary to achieve that goal.

One way to utilize the new creative agent in Creative Cloud is by typing commands in natural language. You might instruct Firefly to examine all of the flyers that you made last summer to promote your swimwear photography business and to change the ethnicity of the model in those flyers. You could then replace the summer scenes in those flyers with winter ones and create text that promotes buying swimwear during wintertime. You could also instruct the program to create variations in appropriate aspect ratios for a variety of social media platforms. You can also have Firefly share to these platforms after you have approved each design. All of these steps and many others can be accomplished simply by speaking or typing a prompt into the creative agent.

As it begins to integrate agentic AI across Creative Cloud, Adobe stresses that the goal is not for you to hand over control of any aspect of the creative process, but rather for you to streamline your workflow so that you spend less time on mundane, repetitive tasks and more time being creative.

Another way to integrate agentic AI into your workflow is to use an Adobe connector, which allows you to access the capabilities of Adobe's tools inside platforms like ChatGPT, Claude, and Slack. Adobe has also previewed local integrations that will allow third-party AI agents to work directly in Photoshop, Illustrator, Premiere, and After Effects, with private beta availability planned for early 2027. In this scenario, you have the option of working primarily through your AI agent or in the Adobe application.

It is also worth noting that you do not have to integrate AI-based tools into your Creative Cloud workflow. There is innovation on the non-AI front as well. Dynamic text additions to Photoshop and Firefly are not AI-based, nor is the new Color Mode workflow in Premiere. Nearly 70% of professional Photoshop users utilize AI-based features regularly, but if you are more comfortable removing objects from your photos using the Clone Stamp tool, there is nothing preventing you from continuing to do so.

If you aren't using these tools regularly, however, be aware that you might be spending time on mundane tasks like saving, moving, and renaming files while your peers are spending that time creating variations of their finished projects.

During the keynote, Adobe's vision for how creatives use the Creative Cloud suite was clear. Apps are designed to perform the same whether you choose to use a tablet, phone, or laptop. Creative tools must be accessible to the user whenever inspiration strikes, no matter what device they have at hand. There is an increasing demand for content each day, and tools added to Creative Cloud must make it easier for professionals to reach, engage, and monetize their audience. Additions to the suite must facilitate collaboration with clients or team members. Collaborators need to be able to navigate shared workspaces intuitively.

Adobe believes the user should always be in control. AI assistants are designed to help get the job done, not to replace the creator’s vision. You are in control of which tasks you perform and which you assign to the creative agent.

Firefly Becomes the AI Creative Studio

Adobe sees its Firefly app as an all-in-one creative studio for brainstorming and preserving ideas. If you were planning a fashion shoot that involved a model, hairstylist, makeup artist, and fashion stylist, you could use Firefly Boards to generate thumbnails that represent the looks that you would like to create for your shoot. You could also import your own photographs to be used as inspiration for your shoot. Each person on the team can have their own board, or they can contribute to others' boards. And because a digital board has the potential to be infinite, you can create boards that feature a lot of different elements or video clips, yet it is still easy to navigate among these elements and understand how they all contribute to the project.

There is a dedicated Firefly Video Editor should you want to animate any elements of the project. Because you aren't leaving Firefly, you don't have to keep track of where project files are located or which version to open in an external program.

Firefly also has several Quick Actions available. These include:

• Upscale image
• Remove object from image
• Expand image
• Remove people

These actions exist so that you can perform multistep operations in Firefly without mastering the program in the traditional sense. A modern mastery of Firefly might require no more than the ability to articulate the end product or result you want from the program. New to the Quick Actions area is a "From Creators" tab. Think of these actions as Lightroom presets, but rather than just changing the look of the image, these actions can help you add a touch of your favorite creator’s style to your own creation.

Firefly Design is a new element that utilizes vector graphics. In a demo provided to me, a flyer was imported into Firefly. Then Firefly was able to deconstruct this flyer as if it had been created in layers. This gives you the ability to modify the font, color, or size and move elements around, just as you could have if you had created this document from scratch using a proper workflow. And because it is vector-based, there is no loss of quality if you enlarge any element of the design.

Motion presets give you the ability to animate text. You can also perform some video editing functions directly in Firefly without opening Premiere or Premiere Mobile. In the demo, I was shown how you could have a still image of something like a food takeout container and give an instruction along the lines of "the camera pans slowly toward the takeout container," and Firefly would create a video from the still image as you described. Creating photographs or video in different aspect ratios for various social media platforms is something that I dislike doing, and I was pleased to see that I can use simple commands like "create social media variations," and the program will create more than ten variations of the same video, photograph, or animated flyer instantly. You can also quickly generate thumbnails for each platform in the correct aspect ratio.

What's New in Photoshop

For a long time, Photoshop was synonymous with Adobe, to the degree that you might hear someone tell you that they edited a photo in Adobe when they meant to say Photoshop. As the capabilities of Lightroom have improved, I find myself rarely using Photoshop. Most recently, I went into the program in an attempt to add a person from one photograph into another. Ultimately, I was unable to do this, and I paid a retoucher to perform the task instead. I believe Photoshop will continue to be used primarily by high-level professionals like the retoucher I sent my job to. Fortunately for the rest of us, many features that have been added to Photoshop over the years have made their way into Lightroom. Among the new additions to Photoshop announced today were Relight, Generative Restyle, and improvements to the Remove tool. Lightroom also received a new Glow feature.

Relight lets you control and reshape the light after the shot has been taken. In Layers, you choose "New Relight Layer." A small icon representing a light source appears on the image. Wherever you move the icon, it appears that light is coming from that source. You can give the light a color to simulate a color gel on that light. You can add up to three lights. Because it's a layer, you can modify it or delete it at any time.

Generative Restyle works from a prompt. In the demo, the prompt was "Turn this scene to golden hour with a pastel sunset sky." The Generative Restyle layer comes with a slider that changes how the effect is applied, so you might get different clouds or different color patterns. You can also type in a number to set the intensity. It bears repeating that you are working in a layer, which gives you full control over how the original image is affected.

I use the Remove tool in Lightroom regularly, and I'm almost always pleased with the result. The Photoshop Remove tool now removes elements that are related to the object that you have removed from the photo. If you select a garbage can, Photoshop now understands that the shadow of that can should be removed as well, even though you did not select the shadow. I haven't had a chance to experiment with this tool yet, but it seems to me that you could just leave the tool set to always remove the related elements when removing an object. This is the type of tool that I can see being added to Lightroom in the near future.

Photoshop also includes a beta feature, similar to the addition in Firefly, that converts flat artwork to text layers. If you collaborate with a teammate who sends you a digital flyer without a source file, you can instruct Photoshop to "convert to text layers." Adobe has 35,000 fonts, so Photoshop can identify the fonts in the flyer or find very close matches. You end up with separate layers for all the text in the design. And it's real vector text, not a picture of text, so you can resize and edit it freely.

As you might expect, Photoshop also has an AI assistant designed to help you perform complicated tasks without having to master the Photoshop interface. One good thing about Adobe adding new features to Photoshop is that it shows a commitment to photography in a world that is increasingly focused on video. Through its integration with Topaz Labs, Photoshop also features improved sharpening, noise reduction, and upscaling of photographs.

During a fireside chat after the keynote, Deepa Subramaniam, Vice President of Product Marketing for Creative Professionals at Adobe, spoke of the importance of spending as much time as possible on creative tasks rather than tedious ones. “Maybe that means going through a few different ideas and conceptualizing it, looking at it critically, and being like, ‘You know what doesn't hit the mark? Let's tweak it this way. Let's tweak it that way.’ That pivoting on an idea or a concept before moving into the actual creative production work just makes the idea even better. It's your taste, your aesthetic, your style that stands out in the world. Spending time on the idea, not the nitty-gritty prep work to get into executing that idea, is a good, worthy exchange of time on your part,” she said.

Resh Sidhu, Vice President of Creative at Adobe, attempted to allay any fears one might have of AI taking over by adding, “Anybody who is in the creative field, anybody who wants to create or make something: the value of the human touch and having an idea is going to remain strong for many decades to come.”

There were many other features announced across the Creative Cloud suite that were not detailed in this article. Premiere now has the ability to separate audio tracks when someone is being interviewed. This is useful when you finish asking a question and your subject answers so quickly that your final word and their first word are merged together. The new feature can isolate your voice from your subject's voice. For more information on this and other features, visit Adobe’s blog. Many of the features described in this article are available today, while others are in beta or preview, with some integrations scheduled for 2027.