# Apple Adds Generative Tools to Photos in iOS 27 > Source: > Published: 2026-06-16 20:19:55.796889+00:00 # Apple Adds Generative Tools to Photos in iOS 27 Apple unveiled new generative features for the **Photos app** in the iOS 27 preview, including **Spatial Reframing**, **Extend**, and an improved **Clean Up**, according to Apple press materials and coverage from Wired and The Verge. Per Apple, developer testing begins immediately and the features will be available to users this fall, according to the Apple press release. Wired reports the tools generate new background pixels but are restricted from altering primary subjects' faces, and that the **Extend** function can expand an image by about **25 percent**. Wired also quotes Apple camera chief Jon McCormack saying these tools give users "superpowers" while stressing a measured approach to AI. ### What happened Apple previewed the next generation of **Apple Intelligence** at WWDC26 and announced new generative editing features for the **Photos app** in iOS 27, according to Apple press releases dated June 8, 2026. The announced photo features include **Spatial Reframing**, which lets users change perspective after capture; **Extend**, which generates new background pixels to add space around a subject; and an improved **Clean Up** tool for object removal, per Apple and coverage in The Verge and MacRumors. Apple says developer testing starts immediately and that these features will be available to users this fall, per the Apple newsroom posts. Wired's hands-on reporting and interviews with Apple staff describe how the generative edits create new pixels limited to backgrounds and that the app prevents altering primary subjects' faces, and Wired notes **Extend** expands images by roughly **25 percent** in early builds. The Verge and Tech outlets provide additional hands-on impressions that the tools generally work but produce mixed results depending on scene complexity. ### Technical details Editorial analysis - technical context: The features combine depth and scene understanding with generative image synthesis, using the phone's depth map to create a rough 3D representation and then filling new regions with model-generated pixels, as described in Wired and PCMag reporting. Industry coverage characterizes **Spatial Reframing** as a fusion of depth-based spatial manipulation and generative inpainting, while **Extend** resembles controlled outpainting constrained to avoid modifying main subjects, according to multiple hands-on writeups. Editorial analysis - technical context: From a practitioner perspective, these capabilities imply tighter integration of on-device perception pipelines with generative models, likely requiring robust depth estimation, semantic masks, and tuned inpainting models to maintain visual coherence at native phone resolutions. Public coverage does not disclose Apple model architectures or exact on-device/off-device inference split; Apple describes the next-generation Apple Intelligence as integrating foundation models across platforms in its press release but does not publish model specs in the announced materials. ### Context and significance The Photos additions place Apple squarely among mobile vendors shipping consumer-grade generative imaging: competitors from Google and Samsung already offer related editing features, and press coverage frames Apple as taking a more conservative, design-led approach to limit edits to backgrounds and protect key subject integrity. Wired quotes Apple camera chief Jon McCormack emphasizing accessibility and deliberate use of AI, while independent reviewers note both impressive results and occasional artifacts. ### Industry context For ML engineers and mobile imaging teams, this release highlights two trends. First, generative features are moving from cloud-first prototypes into mainstream on-device product surfaces. Second, vendors are combining depth maps and segmentation with generative inpainting to reduce visible artifacts. These patterns affect system design choices around model size, latency, privacy, and mixed on-device/cloud inference. ### What to watch For practitioners: Observe how Apple balances model quality, latency, and device privacy. Key indicators include release notes and developer documentation for iOS 27 (Apple Developer notes) that describe APIs and hardware acceleration, the extent of on-device execution versus server-side fallback, and the API surface for developers to interact with Photos editing primitives. Also monitor early adopter reports for common failure modes such as unnatural geometry after reframing, seams at fill boundaries, and performance on lower-end devices. For practitioners: Keep an eye on developer tooling and rate limits in the beta, guidance Apple provides on acceptable uses, and how Apple enforces subject protections in the Photos app versus what third-party apps can do via the App Store. Public reporting indicates Apple restricts edits to background areas in the built-in app, but the broader ecosystem may still expose more permissive generative editing tools. ## Scoring Rationale This is a notable product update for mobile imaging and consumer generative AI, showing mainstream integration of depth-aware inpainting. It matters to practitioners building photo pipelines, on-device inference, and privacy-aware AI, but it is not a frontier research breakthrough. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)