New Photo Editing in iOS 27: Spatial Reframing, Extend and AI Image Labels

iOS 27 brings one of the most substantial changes to photo editing on the iPhone in recent years. Instead of limiting the Photos app to familiar adjustments such as exposure, colour, cropping and object removal, Apple is adding Apple Intelligence tools that can alter the space around a photograph while trying to preserve the character of the original scene. Spatial Reframing can change the apparent viewpoint and composition after a picture has been taken, while Extend can create additional space beyond the existing edges of an image. Apple is also addressing a growing concern around generative editing by marking photographs changed with Apple Intelligence using hidden SynthID information. These additions do not replace traditional photography skills, and they cannot recover genuine visual information that was never captured by the camera. Their main purpose is more practical: correcting awkward framing, adapting a photograph to a different shape, or improving composition without immediately sacrificing important parts of the original frame.

Spatial Reframing Changes the Composition After a Photo Is Taken

Spatial Reframing is designed for photographs where the subject, angle or overall balance of the frame is slightly wrong even though the moment itself is worth keeping. Conventional cropping can only remove parts of an existing picture. It cannot move the apparent viewpoint or reveal additional space beyond the photograph. Spatial Reframing takes a different approach. In the Photos app, the user can interact with the image and change its perspective, with iOS showing a preview of how the scene would look if the camera had been positioned somewhat differently when the photograph was taken. This can make a subject appear better centred, reduce an awkward angle or provide a more balanced amount of space around people and objects. The result is still based on the original photograph, but some portions of the revised frame may need to be generated because the camera never captured them.

Apple says the feature builds on its work with spatial models associated with Apple Vision Pro. For an ordinary iPhone owner, however, the important part is much simpler than the underlying technology. The user does not need to create masks, understand three-dimensional modelling or manually reconstruct a background. Spatial Reframing analyses the scene and generates new visual material only where the change in perspective creates an area that was not present in the original photograph. Areas that do not need to change are intended to remain as close as possible to the captured image. This distinguishes the tool from a complete AI recreation of a photograph. The purpose is to make a limited compositional correction rather than replace the entire scene with a newly generated version.

The most convincing uses are likely to involve moderate corrections. A portrait taken slightly too far to one side can potentially be balanced without cutting away useful background detail. A street photograph may benefit from a small change in viewpoint that makes vertical lines or the position of the main subject feel more natural. Family, pet and travel photographs can also be easier to rescue when the original moment cannot simply be photographed again. Large perspective changes require more invented information and should therefore be treated more cautiously. Faces near the edge of a frame, lettering, repeated architectural patterns, reflections, railings and closely packed objects are particularly worth checking after an AI-assisted edit. Spatial Reframing can make a photograph more usable, but generated areas should not automatically be assumed to represent details that were genuinely present outside the original frame.

How to Use Spatial Reframing Without Losing the Original Character of a Photo

A sensible way to work with Spatial Reframing is to begin with the smallest adjustment that solves the compositional problem. Rather than moving the viewpoint dramatically, the user can first decide what is actually wrong with the picture. The subject may simply need more space in the direction they are facing, a distracting edge may need to disappear, or the horizon and foreground may feel unbalanced. Once the Reframe option is selected in the Photos editor on a supported device, subtle movement can often achieve more believable results than a large transformation. Comparing the revised version with the original photograph is also useful because a technically cleaner composition is not necessarily a better photograph. Expressions, spontaneous gestures and the relationship between people in the frame can matter more than perfect symmetry.

Spatial Reframing should also be considered separately from ordinary cropping. Cropping remains the better choice when everything needed for the final picture is already inside the original image. It introduces no generated scenery and usually gives the photographer more predictable control. Spatial Reframing becomes useful when cropping would remove something important or when the problem involves perspective rather than simply excess space. For example, cutting the left edge from a photograph may centre a person but also remove another person, a landmark or useful environmental detail. A modest spatial adjustment may offer another option. On the other hand, if a photograph contains plenty of unused background, a straightforward crop may still produce the cleaner and more faithful result.

The difference is particularly important for photographs used as records rather than purely personal images. If a picture documents an accident, property condition, news event, product defect or any other situation where exact visual evidence matters, changing perspective with generative tools can alter how the scene is interpreted. In those cases, the untouched original should remain available and the edited version should be identified clearly when it is shared. For holiday photographs, portraits or creative images, the stakes are usually lower, but inspecting the generated areas remains worthwhile. Apple is giving users far more freedom to repair composition after pressing the shutter, yet Spatial Reframing works best when it supports the photograph that was actually taken rather than trying to turn it into a completely different scene.

Extend Adds Space Around a Photograph Instead of Cropping It Away

Extend solves a different problem: a photograph may be good, but its existing boundaries are too tight. Perhaps the top of a building sits uncomfortably close to the edge, a portrait needs additional space above the subject, or straightening a tilted horizon would normally force the Photos app to crop away useful details. Extend can create new image content outside the original edges so that the usable frame becomes larger. Apple specifically presents the feature as a way to give subjects more breathing room, straighten an uneven horizon without sacrificing important content and change the aspect ratio of an image while filling the newly created areas. The feature therefore reverses the normal logic of cropping. Instead of deciding which part of a photograph to discard in order to reach the desired shape, the editor can attempt to add believable surroundings.

This can be particularly useful when the same photograph needs to work in several formats. A tightly composed portrait image may not have enough room to become a wider landscape picture, while a landscape shot can be difficult to turn into a tall image without cutting away the subject. Extend can generate the extra background needed around the existing photograph, potentially reducing the compromises involved in these conversions. It can also help with simpler corrections. Straightening a photograph by several degrees traditionally creates empty triangular areas around the edges, so photo editors automatically zoom and crop the picture. With Extend, those missing edge areas can instead be filled. That can preserve people, objects or scenery that would otherwise disappear after the horizon is corrected.

What Extend creates should still be understood as generated content rather than recovered content. If the original photograph ends at the side of a table, iOS does not know with certainty what was beyond that edge. It can create a continuation that appears visually plausible based on the information already present, but the result is an estimate. Simple skies, walls, grass, water and softly blurred backgrounds are generally easier situations for this type of editing than complicated shop signs, patterned clothing, crowds or objects with precise geometry. Users should therefore inspect the borders of the finished picture carefully. Repeated textures, inconsistent shadows, malformed lettering or objects that change shape near the original boundary can reveal that the expansion has not worked perfectly.

Extend, Crop, Reframe and Clean Up Are Designed for Different Problems

The growing number of editing tools in Photos can make their roles appear similar, but each one answers a different question. Crop asks which parts of the existing frame can be removed. Extend asks what additional space could plausibly be created around it. Spatial Reframing asks whether the apparent viewpoint and composition should change. Clean Up is intended for unwanted objects or distractions already inside the photograph. Choosing between them becomes easier once the problem is identified before editing begins. If a photograph contains too much empty space, Crop is usually enough. If it contains too little space, Extend is the relevant tool. If the position and perspective feel wrong, Reframe is more appropriate. If a distracting object is the main issue, Clean Up addresses that specific element.

Consider a slightly tilted group photograph taken on holiday. Straightening it with an ordinary crop might remove a person’s foot or part of a landmark near the edge. Extend can potentially fill the new gaps created by the rotation and retain the full group. If the main problem is that the group appears pushed too far to one side because of where the photographer was standing, Spatial Reframing may offer a better correction. If a rubbish bin or passing stranger distracts from the group, the enhanced Clean Up tool serves another purpose entirely. More than one tool can be used on the same picture, but applying several generative changes simply because they are available can move the finished image increasingly far from the original moment.

One advantage of editing inside Photos is that users can retain access to the original photograph rather than permanently replacing the captured file with the first edit they make. The Revert to Original option remains important when experimenting with new AI-assisted tools because it provides a straightforward way back if an edit looks unnatural or changes too much. It is also sensible to compare the original and edited versions at a larger viewing size rather than judging only from a small preview. AI-generated edges can look convincing on an iPhone screen while small inconsistencies become more obvious when the photograph is enlarged. For important images, preserving an unmodified copy before extensive editing also makes it easier to distinguish what the camera recorded from what was subsequently generated.

AI Image Labels Make Generative Photo Edits More Traceable

The arrival of more powerful generative editing creates an obvious question: how can someone tell that a photograph has been changed with AI? Apple’s answer in iOS 27 is not a large visible badge printed over the picture. Photos adjusted with Apple Intelligence automatically receive a hidden SynthID watermark that identifies them as having been edited with AI. This distinction matters because the term AI label can suggest that every edited picture will display a permanent warning in a corner. That is not how Apple has described the system. The identification information is intended to travel with the image without covering the photograph itself. It gives compatible systems a way to recognise AI involvement while allowing users to view and share the finished image normally.

Apple has also provided more specific information about editing history inside Photos. During its WWDC 2026 camera and photo technologies discussions, Apple engineers explained that when Spatial Reframe or Clean Up is used, the file’s IPTC metadata is updated to indicate which AI modification was applied. The Photos app can surface that information in the Info panel when the user checks the details of an image. This creates a second useful layer of transparency alongside the hidden SynthID signal. Someone looking at their own library does not have to identify AI editing purely by visual clues. Where the relevant information is available, the photo details can indicate that an AI-assisted modification formed part of the editing process.

These measures improve traceability, but they should not be misunderstood as a universal authenticity test. A photograph without a visible warning should not automatically be treated as completely untouched, because images can be edited in many applications and by many methods that use different systems for recording provenance. Likewise, the presence of an AI marker does not necessarily mean that an entire scene was artificially created. A real photograph may receive a marker because the user extended one edge, removed a distraction or changed its framing. The useful distinction is therefore between the original camera capture and subsequent AI-assisted modification. For journalists, publishers, businesses and anyone working with documentary imagery, keeping the untouched source file and recording what was changed remains a stronger practice than relying on a single indicator alone.

Availability, Supported iPhones and What Users Can Expect in 2026

As of 27 August 2026, iOS 27 has not yet reached its final public release. Apple introduced the software in June, made developer testing available first and subsequently opened public beta testing, with the finished free update scheduled for autumn 2026. This matters when assessing the new Photos features because details can still be refined before the final version ships. Apple has already been correcting issues connected with Spatial Reframing and Extend during beta development. Anyone testing the software early should therefore expect behaviour to change between builds and should avoid treating a beta result as a definitive indication of how the finished editing tools will perform.

There is also an important difference between an iPhone being able to install iOS 27 and being able to use its Apple Intelligence features. Apple lists iOS 27 support for a much wider range of iPhones, extending back to the iPhone 11 family and the second-generation iPhone SE. The Apple Intelligence requirements are considerably stricter. In iOS 27, Apple Intelligence is supported on iPhone 15 Pro and iPhone 15 Pro Max, as well as iPhone 16 models and later generations. Because Spatial Reframing and Extend form part of the Apple Intelligence photo-editing features, owners of older iPhones should not assume that installing iOS 27 will automatically provide these tools. Hardware capability, language and regional availability remain relevant.

For users with a compatible iPhone, the wider significance of the update is not that every imperfect photograph now needs AI correction. The more useful change is that several problems which previously required a separate photo editor can increasingly be handled inside Photos. A slightly poor viewpoint can be adjusted with Spatial Reframing, missing room around the frame can be created with Extend, and distracting objects can be handled by the improved Clean Up tool. At the same time, SynthID and edit information provide a way to distinguish AI-assisted changes from an untouched capture. The best results are likely to come from treating these features as corrective tools: make modest adjustments, inspect generated details carefully, keep the original image and be clear about AI editing whenever the accuracy of the photograph matters.