The Hidden Side of Ingress
How augmented reality games crowdsource data and what that means for privacy in the AI era.
title: "The Hidden Side of Ingress" slug: "hidden-side-ingress" description: "How augmented reality games crowdsource data and what that means for privacy in the AI era." datePublished: "2012-11-28" dateModified: "2026-03-15" category: "AI & Privacy" tags: ["privacy", "AR", "data collection", "surveillance"] tier: 1 originalUrl: "http://www.applieddatalabs.com/content/hidden-side-ingress" waybackUrl: "https://web.archive.org/web/20121128174831/http://www.applieddatalabs.com:80/content/hidden-side-ingress"
The Hidden Side of Ingress
In late 2012, we wrote about a weird little augmented reality game that Google's Niantic Labs had just released into closed beta. It was called Ingress, and the internet was buzzing about it. We weren't interested in the gameplay. We were interested in the data. Specifically, why Google was paying engineers to build a free game that had players walking around photographing landmarks and mapping pedestrian routes with their GPS-enabled phones. The answer, we argued, was that Google was crowdsourcing the largest pedestrian mapping dataset ever assembled, and nobody playing the game seemed to care.
We were right. And what happened next was bigger than any of us expected.
What We Said in 2012
Our original analysis was pretty straightforward. Google had a history of building free products that were really data collection tools in disguise. We pointed to GOOG-411, the free directory assistance service that seemed pointless until you realized Google was using millions of voice queries to train its speech recognition engine for Android. Ingress followed the same playbook.
The game required players to walk around with GPS running, take geo-tagged photos of real-world landmarks, and upload everything to Google's servers. "Through this process, Google is collecting vast amounts of pedestrian data, average walking speed, routes taken, pictures, and even Wi-Fi hotspots available," we wrote. Google Maps didn't handle pedestrian navigation well at the time, and Nokia had just announced pedestrian turn-by-turn directions. We predicted Google would use Ingress data to build the best pedestrian maps available.
The players themselves didn't mind. One user we quoted said: "Very fun to play. I don't mind allowing them to gather data, with the idea that I will need to use the service that will be developed from it." That sentiment turned out to be prophetic for the entire decade that followed.
Pokemon Go Proved the Thesis, Then Some
In July 2016, Niantic spun out of Google and launched Pokemon Go. The game hit 500 million downloads in its first two months and generated over $1 billion in revenue by the end of that year. By 2023, lifetime revenue had surpassed $6 billion. Niantic CEO John Hanke had been running the Ingress data collection experiment the whole time, and Pokemon Go was the scaled-up version.
Google didn't build Ingress to make a game. They built it to make every phone in the world a mapping sensor. Pokemon Go proved the model could work at a scale nobody imagined.
The data collection was staggering. Niantic gathered detailed 3D mapping data from millions of players who scanned real-world locations using their phone cameras through a feature called "Niantic Lightship." This wasn't just pedestrian path data anymore. Players were creating detailed spatial maps of parks, plazas, storefronts, and public spaces worldwide. Niantic explicitly built a "Visual Positioning System" from this crowdsourced data and licensed it to other companies.
But Pokemon Go was just the beginning of consumer-product-as-data-vacuum.
The AR Data Pipeline Goes Mainstream
Apple launched Vision Pro in February 2024 at $3,499, and while sales were modest, it contained LiDAR sensors and cameras that continuously scanned the user's environment to create spatial maps of rooms, furniture, and objects. Meta shipped its Ray-Ban smart glasses in October 2023, and by mid-2025, they'd sold over 10 million units. Those glasses have cameras pointing at everything the wearer sees. Meta's stated purpose is to build AI assistants, but the environmental data those cameras collect is a spatial computing goldmine.
The pattern we identified in 2012 has become the default business model for consumer tech. Give away (or sell at a loss) a device or app that captures environmental data. Use that data to build mapping, spatial computing, and AI training datasets. Monetize the data through services, licensing, or ad targeting. Google, Apple, Meta, and Niantic are all running this play simultaneously.
What's changed since 2012 is the sophistication. Ingress collected GPS coordinates and photos. Today's devices collect 3D spatial maps, gaze tracking, hand gestures, voice data, and continuous environmental audio. Meta's smart glasses can identify objects and read text in real time. Apple's Vision Pro tracks where you look with millimeter precision. The amount of personal and environmental data flowing through these devices dwarfs what Ingress collected by orders of magnitude.
What This Means for Enterprise Data Governance
For enterprises, the implications are serious. Employees are walking into offices, factories, and sensitive facilities wearing devices that continuously capture spatial and visual data. A worker wearing Meta Ray-Ban glasses in a pharmaceutical lab is potentially streaming proprietary equipment layouts to Meta's servers. A visitor wearing Apple Vision Pro is mapping your entire office with LiDAR.
This is exactly the kind of problem that enterprise data governance frameworks need to address. It's not enough to have policies about what data your company collects. You need policies about what data your employees' consumer devices are collecting while they're on your premises and sending to third parties.
The companies that get this right are building what we'd call AI-ready infrastructure -- they're thinking about data flows not just inside their systems, but in and out of their physical spaces. And as Operational AI adoption accelerates, the line between "consumer data collection" and "enterprise data exposure" will keep blurring.
The Trade-off We Accepted
Back in 2012, we ended our original piece by noting that Ingress users didn't mind the data collection because the game was fun. Fourteen years later, that bargain has scaled to billions of people and trillions of data points. We've collectively decided that free maps, fun games, and AI assistants are worth the continuous surveillance of our physical world. That Ingress user who said "I don't mind allowing them to gather data" was speaking for an entire generation, whether they knew it or not.
The question for the next decade isn't whether consumer devices will collect spatial data. They will. The question is who owns that data, who can access it, and what happens when the 3D map of your office that an employee's smart glasses captured shows up in a competitor's AI training set.