Want to Be Completely Unsurprised? Pokémon Go Data Helped Train Navigation Systems for Military Drones

If you played Pokémon Go in 2016 (or later), congratulations — you may have unwittingly helped train AI for military drones.
A Dutch investigation by the newspaper Trouw has revealed how Niantic turned billions of player-generated environment scans into valuable geospatial training data. The company’s subsidiary, Niantic Spatial, used this data to build a Visual Positioning System (VPS) that allows robots and drones to navigate without relying on GPS.
Also read:
- Sony Invests $100 Million in Cinemas — But Not the Cinemas You’re Thinking Of
- One in Three TikToks Shown to New Users Is Low-Quality AI Slop, Study Finds
- Who’s to Blame When an AI Agent Messes Up? The Accountability Gap in Agentic AI
- Remember That Study That Scared AI Fans a Year Ago? It Just Spectacularly Fell Apart.
The Scale Is Mind-Boggling

The technology, called VPS, enables machines to understand and locate themselves in the real world by matching camera input against a massive pre-built spatial database. It’s particularly useful in GPS-denied environments — think dense urban areas or active conflict zones.
From Catch ’Em All to Defense Tech

In today’s world, the most in-demand “smart robots” are often the ones operating in contested environments. Navigation systems that work when satellites are jammed or unavailable are extremely valuable.
The 2016 Data Blind Spot

Personal data awareness was also far lower then. Today we’re much more cautious about what apps can see and record. In 2016? Not so much.
(Note: The author of this piece did not play Pokémon Go and, at the time, strongly criticized those who did.)
The New Normal
Niantic has since spun off its games business and kept the spatial technology platform. The company maintains it has the right to commercialize the data under the terms players agreed to. Technically, they’re correct — users opted in (often for minor rewards).
This story is a near-perfect case study in how data collected for one harmless-sounding purpose can find its way into high-stakes applications years later. What seemed like innocent fun in 2016 became raw material for tomorrow’s autonomous systems.
Want to be completely unsurprised? This is exactly how it was always going to end. The only surprise is how few people saw it coming at the time.
Subscribe to our newsletter
Get the latest Web3, AI, and crypto news delivered straight to your inbox.