What Apple’s New AI Photo Editing on iPhone Actually Does
Apple’s new AI photo editing suite in the iOS 27 Photos app combines Spatial Reframing, generative Extend, improved Cleanup, and deeper RAW image processing so users can reshape composition, fix distractions, and refine detail after capture, by generating synthetic pixels only where needed while preserving original sensor data for more authentic and controllable edits overall. In iOS 27, Photos stops being a light retouching tool and starts to resemble a consumer-friendly pro editor. Apple Intelligence can expand a scene beyond the original frame, shift perspective as if the camera had been moved, and remove people or clutter that spoil a shot. At the same time, a major update to Core Image RAW improves denoising and color for apps that work with RAW files. The result is a stack of tools that help you rescue imperfect photos while forcing a fresh conversation about what still counts as a faithful image.
Spatial Reframing: Moving the Camera After the Fact
Spatial Reframing is the headline AI photo editing iPhone feature, because it tries to recreate a shot you never took. Instead of cropping, you drag the frame around and the scene appears to have been captured from a different position. Apple combines on-device spatial models, informed by Vision Pro work, with image generation on Private Cloud Compute to estimate the 3D layout and fill in missing areas. According to Digital Trends, Apple says only the newly exposed portions of the image are generated by AI, while the original content remains untouched. In practice, you might lower a family photo to improve headroom or shift sideways to center a landmark, and the Spatial Reframing tool invents background around the subject. Early demos look natural, but testers have already seen occasional distorted faces, so it is powerful but not flawless generative photo editing.
Extend and Cleanup: Fixing Tight Framing and Distractions
Alongside Spatial Reframing, the iOS 27 Photos app gains two practical Apple Intelligence tools: Extend and an upgraded Cleanup. Extend tackles the classic too-tight smartphone composition by expanding a photo outward instead of cropping inward, generating new pixels around the original frame. You can turn a cramped vacation shot into a wider scene or reshape an image to fit a wallpaper or social post without losing your subject. Cleanup focuses on removing distractions: passersby in travel photos, clutter behind portraits, or unwanted props in product shots. Earlier versions worked best with small objects and simple backgrounds; the new Cleanup reconstructs larger areas more convincingly. You can pick between Fast, High Quality, or Auto models depending on whether speed or fidelity matters more. These additions make the iOS 27 Photos app a far stronger everyday editor, reducing the need to open a separate app when a frame or background spoils an otherwise good image.
RAW Image Processing Grows Up After a Decade
Away from the Apple Intelligence branding, RAW image processing quietly receives its most meaningful overhaul in years. Apple is rolling out version 9 of its Core Image RAW engine across iOS 27, iPadOS 27, and macOS Golden Gate. This update improves denoising, color correction, and overall sharpness based on the original sensor data, giving photographers more headroom before they even start editing. This change lives under the hood rather than as a toggle in the default iOS 27 Photos app. It is aimed at third-party camera and editing apps such as Halide or Pixelmator Pro, which can now draw on cleaner, more flexible RAW files. Importantly, Apple Insider notes that this new RAW stack is not directly tied to extending or reframing features. That separation underlines Apple’s dual track: computational photography and generative photo editing on one side, and more traditional RAW workflows that prioritize fidelity on the other.
Authenticity, Ethics, and Apple’s ‘Not AI for AI’s Sake’ Pitch
Together, Spatial Reframing, Extend, Cleanup, and the RAW upgrade turn the iOS 27 Photos app into a powerful, generative editing hub. Yet they also blur the line between fixing a shot and fabricating a new one, because many results now depend on synthetic pixels. Crops, color tweaks, and basic noise reduction were always interpretive; moving the apparent camera position and inventing new scenery pushes that boundary much further. Apple’s camera leadership has stressed that the goal is not to add AI for the sake of AI, but to solve familiar photo problems: cramped framing, awkward perspective, distracting background elements, and flat-looking RAW files. For casual users, that means more shareable images with less effort. For professionals and journalists, it raises the need to separate documentary images from AI-modified ones. Expect future debates about labelling, workflow discipline, and when an edited photo stops being a photograph and becomes an illustration.
