iOS 27 is turning the Photos app into something that feels less like a basic editor and more like a quiet, on-device retoucher who understands what you were trying to capture when you hit the shutter.

Apple is rolling out three new Apple Intelligence-powered tools in Photos – an upgraded Clean Up, a new Extend feature, and a Spatial Reframing tool – and together they change not just how you edit photos, but how confident you can be about a shot after it is already taken.

For years, Photos has been a “good enough” editor for casual tweaks: a bit of exposure here, some vibrance there, and the occasional crop when your composition was slightly off. If you wanted anything more ambitious – removing an awkward stranger, widening a frame for a wallpaper, or faking a slightly different camera angle – you typically had to jump into something like Photoshop, Lightroom, or a dedicated AI editor. With iOS 27, Apple clearly wants to close that gap by baking generative AI directly into the default Photos experience, but in a way that still feels “Apple” – opinionated, privacy-conscious, and fairly restrained compared to the wild west of some third-party tools.

The first piece of this trio is Clean Up, and on paper, it sounds like a simple upgrade to a feature that has technically been around for a while. Previously, Clean Up was Apple’s answer to Google’s Magic Eraser or the classic content-aware fill – highlight a distraction in the frame and the system tries to remove it and guess what should go in its place. The issue was that it worked great for basic, low-contrast backgrounds, but it often struggled when scenes got complicated: busy city streets, overlapping subjects, or anything with repeated patterns like fences or railings.

In iOS 27, Apple is calling this a “big upgrade,” and early hands-on reports back that up. Clean Up is now powered by Apple Intelligence, which lets the system build a more detailed understanding of the scene before it starts painting pixels back in. When you brush over a photobomber, a stray trash can, or that one power line slicing through your sunset, the infill does a better job of matching textures, lighting, and even subtle shadows, making the edit less obvious unless someone is zooming in to hunt for seams.

There is also a bit more control baked in. Under the hood, Apple lets Clean Up choose between different processing profiles – a “fast” mode that prioritizes speed, a “high quality” mode that takes a bit longer but does more detailed synthesis, and an automatic setting that aims to balance both based on the scene. That is the kind of small detail that matters in day to day use: sometimes you just want to quickly erase a signboard before posting to Instagram, other times you are cleaning up a shot you plan to print.

If Clean Up is Apple’s upgraded eraser, Extend is its answer to a very specific creative regret: “I wish I had stepped back just a little.” At a high level, Extend is Apple’s take on Adobe’s Generative Expand and Google’s similar “expand the photo” tricks on Pixel phones – you can effectively pull back from your original frame and let AI hallucinate what would have been just outside the edges of your shot.

The implementation in iOS 27 lives under a new Tools section in the Photos editor marked with the Apple Intelligence logo. You select a photo, tap Edit, then tap Tools, and from there you can choose Extend to start widening the canvas. As you drag to change the aspect ratio or give your subject more breathing space, the app shows a blurred overlay where new content will be generated, making it clear which parts of the frame are real and which are AI.

Apple is pitching Extend as a practical tool, not just a party trick. In its demos, the company showed common scenarios: straightening a crooked horizon without having to crop out parts of the frame, preparing a landscape for a desktop wallpaper by adding more sky, or reframing a portrait for print by extending the background instead of chopping off someone’s arm. Because the generation is guided by Apple Intelligence and tuned for realism, backgrounds it synthesizes tend to mimic existing patterns and lighting – more “plausible extrapolation” than wild reinterpretation.

That said, Extend still falls into the generative AI territory where the line between document and creation blurs. When the tool is filling in a bit of extra grass at the edge of a field, most people will see that as harmless; when it is constructing more of a building, or guessing what was behind someone’s shoulder, you start to move into a space where ethics and expectations become fuzzier. Apple’s positioning so far emphasizes “respecting the original moment,” which is a polite way of saying that the system tries to stick close to what is already in the frame rather than inventing dramatically new elements.

The third feature, Spatial Reframing – simply called Reframe in the app – is the one that really separates Apple’s approach from most of the competition. On the surface, it sounds almost impossible: change the camera angle after you have taken the photo. Under the hood, it is powered by the same kind of spatial modeling that Apple has been building for years for features like Portrait mode and, more recently, visionOS and Apple Vision Pro’s 3D imagery.

Here is how it works in practice. When you pick Reframe, Apple Intelligence takes a moment to analyze the photo and construct a sort of spatial map of the scene – an estimation of where subjects and objects sit in three dimensional space. Once that map is ready, you can drag your finger to subtly shift the perspective: tilt left or right, nudge up or down, or even make small changes that mimic stepping a bit closer or farther away. As with Extend, parts of the new frame that do not exist in the original image are filled in on the fly by AI, but the goal is to feel like a small composition fix rather than a full synthetic scene.

This feature builds on Apple’s existing “spatial” photos and videos, but Apple says it works on regular photos as well, including shots taken on non-Apple cameras that you have imported into your library. Imagine a group photo where the main subject ended up slightly off center or an otherwise perfect candid where eye contact is just a fraction off; Spatial Reframing is designed to let you correct those kinds of near misses without reshooting. It is also the most ambitious – and likely the most controversial – of the trio, because it comes closest to rewriting what the camera saw in the moment.

All three tools – Clean Up, Extend, and Spatial Reframing – live under that Apple Intelligence “Tools” panel in the iOS 27 Photos app. They run using a blend of on device processing and Apple’s Private Cloud Compute infrastructure, meaning heavier lifting can be offloaded to Apple’s servers while the company insists that user photos remain protected and are not used to train broader models. For Apple, that privacy story is not just a talking point; it is a differentiator in an AI space where many rivals lean heavily on constantly streaming data back to the cloud.

Timing wise, these tools are already present in the iOS 27 developer beta, with a public beta expected later in the summer and a general release later in the year, aligning with new iPhone launches. They also sit alongside broader Apple Intelligence capabilities – natural language search in Photos, AI generated memory movies, and smarter on device understanding of your library – that aim to make both finding and editing photos feel more conversational. For users, that translates to a Photos app that can help you locate “that picture where I’m wearing a green hoodie at the beach” and then let you fix the horizon, remove a stranger, and widen the frame, all without leaving the default editor.

Of course, the bigger question is what this means for photography culture in general. As AI editing becomes this accessible, it is no longer just professional photographers or power users who can reshape their images in sophisticated ways. The average person can now take a cluttered, slightly crooked shot in a crowded place and, with a few taps, turn it into what looks like a carefully composed, distraction free image where they just happened to be in the right place at the right time.

Apple seems keenly aware of the tension. The messaging around these tools is deliberately conservative – “enhance their images in ways that respect the original moment” – and the interfaces push you toward small, plausible adjustments rather than extreme transformations. Compared to some third-party apps that proudly advertise wild AI remixes and fantasy backgrounds, Photos in iOS 27 feels more like a precision scalpel than a magic wand.

For creators, journalists, and anyone who lives online, that subtlety is going to matter. When tools like Spatial Reframing and Extend are this easy to use, it will be even more important to be clear about what has been edited, especially in contexts like news, reviews, or documentary work where audiences reasonably expect authenticity. At the same time, it is hard not to appreciate the practical benefits for everyday photography: the family photo rescued from a weird crop, the travel shot saved from a rogue tourist, the skyline straightened without sacrificing your composition.

In a way, iOS 27’s Photos update is Apple’s clearest statement yet about how it wants AI to show up in consumer products. The company is not chasing the most dramatic generative demos or turning the Photos app into a blank canvas for surreal edits; instead, it is using AI to gently close the gap between what you saw, what you meant to capture, and what ended up in your camera roll. Whether you are a casual iPhone user or someone who lives in their photo library for work, these three new tools are going to quietly change how much you trust that “one quick shot” – because now, you have a lot more room to fix it after the fact.