Apple 在 2026 年 6 月 8 日的 WWDC 宣布 iOS 27 将于秋季更新 iPhone Camera app，核心变化是把 Siri AI 直接放入相机介面，而不只是 Photo 与 Video 分页。过去使用者必须按 Camera Control button 才能启动 Siri 的 Visual Intelligence；现在功能可见性提高，能以类似 Google Lens 的方式辨识镜头前内容，例如把演唱会海报日期加入日历或辨识植物。
这次更新也显示 Apple 正把 computer-vision AI 作为跨装置能力铺路。文章指出，这类视觉 AI 协助已是 smart glasses 与带相机 AI pendants 的关键功能，因为使用者可单次点按取得协助，不必输入或解释。Apple 据报也在研发 smart glasses 与内建相机的 AirPods，因此 iPhone 上的首次亮相可能延伸至其他产品。
Photos app 将新增或强化 3 项 AI 编修工具：Clean Up、Extend、Spatial Reframe。Clean Up 已存在，但因更新的 AI foundation models，移除物件后的补图更真实。Extend 用 generative AI 扩展照片边界，补出它推测应存在的画面，功能接近 Google Pixel 近几年的工具。Spatial Reframe 可用手指拖曳稍微改变照片视角，并以 generative AI 填补新视角造成的空缺，但是否限于特定 iPhone 仍未明。
Apple announced at WWDC on June 8, 2026, that iOS 27 will update the iPhone Camera app this fall, with the central change being Siri AI placed directly inside the camera interface rather than only Photo and Video tabs. Previously, users had to press the Camera Control button to trigger Siri’s Visual Intelligence; now the feature is more visible and can identify what the camera sees in a Google Lens-like way, such as adding concert-poster dates to a calendar or identifying a plant.
The update also shows Apple positioning computer-vision AI as a cross-device capability. The article notes that this kind of visual AI assistance is already a key feature of smart glasses and camera-equipped AI pendants, because users can get help with a single tap without typing or explaining. Apple is reportedly developing smart glasses and AirPods with embedded cameras, so the iPhone debut may later extend to other products.
The Photos app will add or improve three AI editing tools: Clean Up, Extend, and Spatial Reframe. Clean Up already exists, but updated AI foundation models should make post-removal image filling more realistic. Extend uses generative AI to expand photo edges and create what it predicts should be in the scene, closely matching Google Pixel tools from recent years. Spatial Reframe lets users drag a finger to slightly change a photo’s perspective and uses generative AI to fill gaps, though device limits remain unclear. (Key numbers: 3)