NearLens was built to answer a real privacy problem: anyone nearby might be recording with camera-equipped smart glasses, and bystanders have no way to know. The challenge was to use AI to build a functional product that addresses this gap. The core problem is asymmetry: glasses never announce filming, facial recognition can identify strangers from a single frame, and bystanders lacked an independent signal before capture. Seven smart-glasses owners were interviewed across Waterloo, Atlanta, San Francisco, Indianapolis, Ottawa, and Toronto. Key insights: glasses hide filming by design, none of the seven read the terms in full, detection split reactions 3 positive, 2 against, 2 conditional, and the microphone worried interviewees as much as the camera. The product architecture runs entirely on the user's phone in three steps: listen (passive BLE scan), match (company IDs checked against known list: Meta, Snap, Samsung, Google), and alert (haptic notification, log written locally, no server, nothing transmitted). Field testing across four locations found 15 likely smart-glasses detections among roughly 7,000 people at TD Place Arena Ottawa, 0 detections among 823 Bluetooth devices at Meridian Theatres Ottawa, and single detections at airports. Privacy Intelligence is available in the current NearLens release. The base app can be installed on iOS 15.1. Natural-language Q&A requires iOS 26, Apple Intelligence enabled and available, and a compatible iPhone: iPhone 15 Pro models or iPhone 16 or later. The app ships with WCAG 2.2 AA accessibility from the first commit: VoiceOver labels, dynamic announcements, reduced-motion support, and haptic feedback. NearLens is free on the App Store with no subscription, no advertising, and no data collected.
