Adobe’s Project Indigo Adds Generative AI Tools in a Test of the App’s Future

Updated July 20, 2026 7:28 pm

Adobe has begun testing generative AI features inside Project Indigo, its iPhone camera app originally pitched as a way to deliver a more natural, SLR-like look. The update adds AI-powered filters, object removal, photo advice, and prompt-based editing, marking a major shift for an app that launched as a photography tool first and an AI playground second.

The experiment matters because Adobe is now blending computational photography with generative editing in a mobile camera app, while also testing how much AI users actually want in a product that once appealed to purists seeking realism rather than heavy manipulation.

What changed in Project Indigo’s latest update?

Adobe’s version 1.1 update for Project Indigo introduces a new “AI Playground” that packages several generative tools into the camera app. The company says the feature set is being rolled out experimentally to a small share of users over the next few weeks, and access does not require users to sign in.

Importantly, Adobe says the AI module is optional. A toggle allows users to decline the new system and keep using Indigo in its earlier form, which focused on manual camera controls and image quality rather than AI-driven alteration.

The update contains four major functions: Styles, Object Editing, Photo Guidance, and Custom Edit. Adobe fellow Marc Levoy said the tools rely on engineered prompts rather than open-ended prompting alone, and he noted that repeated uses may produce slightly different outputs.

AI Playground features at a glance

Why is Adobe adding generative AI to a camera app built for realism?

Adobe is doing it because the company appears to be testing a broader vision for mobile creation: taking a photo, improving it, and editing it within one app using both computational photography and generative AI. That is a meaningful departure from Indigo’s original identity, which was built around natural-looking iPhone photos and manual control.

When the app launched, it was described as a tool for camera enthusiasts who wanted a more faithful photographic experience, not a synthetic one. The new update suggests Adobe sees a future in which those two approaches can coexist inside the same workflow.

Levoy said Adobe sees this as the start of a larger effort to build an integrated mobile camera-and-editing experience that combines modern computational photography with AI.

The company has not said the current experiment is permanent. Instead, it is actively studying how people use the tools before deciding whether to extend the test, expand access, or release a paid version later.

How the new AI tools work

The AI Playground is designed to make image manipulation more accessible without requiring users to master separate editing software. Each section in the update serves a distinct role, from quick visual effects to more detailed text-based adjustments.

Styles creates instant AI looks

Styles applies preset generative effects to photographs taken in Indigo. Adobe says the results can alter the mood, lighting, and even the apparent medium, making a picture resemble an illustration or a heavily stylized image.

Unlike a traditional filter, these looks are not simple overlays. They are driven by generative AI and can therefore produce different outcomes when triggered again.

Object Editing removes distractions

Object Editing is intended to delete unwanted people, clutter, or background interruptions. After removal, generative AI reconstructs the missing area so the final image appears complete.

For many users, this will look familiar. The feature sits in the same product category as Google’s Magic Eraser and other AI cleanup tools now common across consumer photo apps.

Photo Guidance offers editing advice

Photo Guidance is less about direct manipulation and more about feedback. Adobe says the tool looks at an image and suggests ways to improve the shot, whether by taking another photo or adjusting the current one.

That makes the feature more like an AI-assisted photo tutor than a conventional editor. For casual users, it could reduce guesswork when deciding whether a shot is worth keeping or reshooting.

Custom Edit lets users describe what they want

Custom Edit is the most flexible part of the new suite. It allows users to describe in more detail how they want an image changed, which can be useful if a preset effect goes too far or misses the intended style.

Adobe says it is using Google’s Nano Banana model for the feature set to start, though the company has left the door open to swapping models or adding support for others later.

What makes this update different from Adobe’s usual AI strategy?

The most notable difference is that Project Indigo’s AI experiment does not yet carry the same level of transparency messaging Adobe has typically attached to its Firefly ecosystem. Adobe’s generative AI products often emphasize content authenticity and clear labeling, but the Indigo update is still evolving in that regard.

Levoy said the team is working on support for C2PA Content Credentials, a metadata standard meant to help identify AI-generated or AI-modified content. He also noted that Google’s Nano Banana model already uses SynthID, an invisible watermarking system.

That means the app is moving into a sensitive area: image generation and alteration in a camera product that originally aimed to preserve a more authentic look. The tension between convenience, creativity, and disclosure is now part of the experiment.

Levoy argued that whether an edit is acceptable or misleading depends on the context, including how the image is shared and presented, and he stressed that technology alone cannot replace responsible behavior.

How did Project Indigo start?

Project Indigo began as a niche camera app for iPhone users who wanted better-looking mobile photos without the overprocessed look that often comes from smartphone pipelines. It combined computational photography techniques with manual settings such as focus, shutter speed, ISO, and white balance.

That launch positioning gave the app a distinct identity. It was aimed at users who care about image quality, not just fast social posting, and it was widely seen as a tool for enthusiasts rather than the mainstream casual market.

The new AI layer does not erase that origin, but it does complicate it. The app is shifting from being primarily a capture tool to becoming both a capture and transformation platform.

From camera control to image generation

Adobe’s direction suggests a broader industry trend: camera apps are no longer just about recording reality more accurately. They are increasingly becoming interfaces for reshaping it immediately after capture.

That shift has obvious consumer appeal. It also raises questions about whether users still want separate stages for shooting and editing, or whether they are comfortable letting AI intervene from the moment a photo is taken.

Why this matters for mobile photography and AI editing

The move is significant because Adobe is testing the idea that generative AI can live inside a photographer-focused app without destroying its appeal to serious users. If the experiment works, it could influence how other camera and editing products are built.

There is also a commercial angle. Adobe has indicated that if the feature set proves popular, a paid version could follow. That would make Indigo not just a product experiment, but potentially a new subscription or upsell path for the company.

For Adobe, the test also helps answer a broader strategic question: can an app founded on realism attract users to AI-powered enhancement without losing the trust that made it interesting in the first place?

• It blends capture and editing into one workflow.
• It tests demand for generative tools in a photography app.
• It may become a paid feature if users respond positively.
• It raises transparency questions around AI-modified images.

What role is Google playing here?

Google is supplying the model behind at least part of the experiment. Adobe says it is starting with Nano Banana, which handles some of the generative work inside the new AI Playground.

This is notable because Adobe is not relying solely on its own Firefly stack for the update. Instead, it is using an external model to move quickly and to test which capabilities users value most.

Google’s involvement also highlights a larger trend in consumer AI products: model flexibility is becoming a competitive advantage. App makers increasingly want the option to swap providers rather than build every feature around a single in-house system.

How the rollout is being tested

Adobe says only a small percentage of Indigo users will see the feature during the current trial period. The company is using the test to study behavior before deciding whether to widen access or introduce more experiments later.

That measured approach suggests Adobe does not want to force a major identity change on the app all at once. Instead, it is probing whether users embrace AI-assisted editing as a natural extension of the camera experience or reject it as a distraction from Indigo’s original purpose.

Current test conditions

What happens next?

The next phase will likely depend on user reaction. If people find the AI tools helpful, Adobe could expand the test and add more models or more polished versions of the features. If users prefer Indigo in its original form, the company may keep the AI layer more limited or optional.

Adobe is also likely to refine its transparency and provenance features as the experiment continues. Content credentials and invisible watermarking are becoming increasingly important as companies try to keep AI-generated and edited media distinguishable from untouched photography.

For now, the update is best understood as a live product experiment rather than a finished pivot. Adobe is watching to see whether a camera app that began with a purist philosophy can evolve into something broader without losing the audience it first attracted.

The bigger picture

Project Indigo’s new AI Playground is more than a feature update. It is a signal that the line between camera, editor, and generative AI tool is fading quickly.

That may be welcome news for people who want faster edits, smarter corrections, and more creative control on a phone. It may also unsettle users who prefer a clearer boundary between photography and synthetic image generation.

Either way, Adobe is now testing the future of mobile photography in public. The outcome could shape not only Indigo, but also how the next generation of camera apps decide whether to preserve reality, enhance it, or reinvent it on the fly.

Adobe’s position, as described by Levoy, is that the company is still learning how people want to use generative editing and will adjust the product based on the results.

That makes this less a finished product launch than a referendum on how much AI belongs inside the camera itself.