Apple’s iOS 27 developer beta introduces the first suite of substantial generative AI editing tools for the Photos app, marking a significant shift in the company’s approach to native image manipulation. The update features three primary capabilities: Clean Up, Extend, and Spatial Reframing. While Apple’s implementation remains more restrained than Android competitors, the tools represent a functional milestone for cloud-hybrid editing workflows. Clean Up has received a fundamental overhaul. Moving beyond the limited on-device processing of previous iterations, the updated tool now leverages cloud-based generative models to remove unwanted objects and reconstruct backgrounds with significantly higher fidelity. Testing confirms the feature reliably eliminates distractions while maintaining visual continuity, addressing long-standing complaints regarding artifacts in prior versions. Extend functions as a reverse-crop mechanism, utilizing AI to expand image boundaries and add plausible background elements. The tool operates with intentional constraints, prioritizing symmetry and avoiding direct modification of human subjects. In practice, Extend provides subtle compositional padding that generally blends seamlessly with existing lighting and geometry, though it occasionally introduces unverifiable elements such as background foliage. Spatial Reframing presents the most technically ambitious but also the most problematic feature. The tool simulates three-dimensional camera movement to recompose existing two-dimensional images. Adjustments remain limited to the approximate physical range of a photographer’s arm. While effective for minor perspective corrections, the feature struggles with close-range subjects. Recombining facial details often produces uncanny distortions, and recomposing crowded scenes frequently results in algorithmically generated figures that lack structural accuracy. Apple has implemented Synth ID labels to flag images modified by these tools. Social media platforms recognize these metadata tags, though visibility is restricted to explicit user queries. The labeling system acknowledges growing concerns regarding digital authenticity. As generative editing becomes increasingly accessible, the distinction between documented photography and algorithmically reconstructed imagery continues to erode. Minor manipulations that introduce unverifiable elements or alter spatial relationships contribute to a broader credibility gap in user-generated content. These features are currently available only to developers and will undergo further refinement before a public release. Apple’s conservative implementation strategy prioritizes reliability over creative control, positioning iOS as a platform for practical photo enhancement rather than aggressive generative manipulation. The success of these tools will ultimately depend on public acceptance of their limitations and the continued evolution of metadata standards for AI-edited media.