As technology evolves, the definition of a photograph becomes increasingly unclear. With big companies incorporating generative AI capabilities into their devices, the distinction between real images and artificially altered ones continues to dissolve. Google and Samsung have already introduced features that enable users to erase or reposition people in photos and even insert new objects into scenes.
Apple is also entering this landscape with innovative features in its forthcoming iOS 27 Photos app. Jon McCormack, Apple’s camera chief, asserts that the company is not employing AI merely for its own sake. Instead, they aim for a more thoughtful integration compared to competitors.
During the recent Worldwide Developer Conference, Apple showcased several enhancements to the Photos app. These include access to improved AI models, enhancing the existing Clean Up tool, which erases unwanted elements in images. The new features—Extend and Spatial Reframe—allow users to expand the surrounding space of a photograph or change its perspective while generating artificial pixels to fill in the gaps.
McCormack highlights that these advancements are carefully designed to empower everyday users with “superpowers” that simplify complex editing tasks. For instance, Extend generates pixels based on what it presumes should be in the enhanced space, although it deliberately avoids altering the primary subject within the image. Additionally, to ensure authenticity, generated pixels stay in the background, preventing the alteration of the main subjects. The Extend feature provides only a one-time expansion, up to 25%, and cannot be repeatedly applied.
In terms of preserving the authenticity of images, Apple plans to implement Google DeepMind’s SynthID technology in the future. This will add an invisible watermark to indicate that a photo has been altered using generative AI. Platforms where photos are shared might flag these images as AI-edited.
McCormack maintains that a photograph represents a real event, underscoring the company’s commitment to maintaining the integrity of captured moments. Conversely, Google permits users to significantly manipulate images to better reflect their memories.
Apple’s approach is meant more for correcting compositional oversights rather than allowing expansive creative alterations. For example, the AI helps to minimize distractions like an unwanted object or adjusts the framing of a subject that is too close to the edge. Della Huff, product manager for Apple’s Camera and Photos software, emphasizes the focus on ensuring that the AI doesn’t generate unnecessary elements in images.
However, there are concerns about how AI models perform in practice. During testing, extending a photo can lead to unexpected results, such as the inclusion of nonexistent people in the generated background. While Apple is refining its AI, it acknowledges that sometimes, creating plausible continuity is necessary for usability.
With these advancements, Apple plans to set limits on the use of these features, emphasizing a structured approach rather than unregulated creative freedom. To use features like Extend, Spatial Reframe, or Clean Up multiple times per day, users will need an iCloud subscription.
A major theme of Apple’s developer conference was interactivity. Siri will now play a role within the Camera app, allowing users to utilize its capabilities without heavy command protocols. However, the editing features will remain user-operated rather than entirely controlled by Siri, due to the complexities involved in such tasks.
In summary, while Apple is introducing new generative AI features aimed at empowering users to improve their photography experience, it is also navigating the challenges of maintaining image authenticity and user experience. Apple aspires to balance creativity with the sanctity of real memories captured in photographs.
