Contents
From Mobile Gaming to Military Intelligence: The Unseen Journey of Your Pokémon Go Data
For years, Pokémon Go players eagerly scanned streets, parks, and landmarks, exchanging real-world augmented reality (AR) scans for in-game rewards. Unbeknownst to many, these playful expeditions would yield a trove of data that could potentially fuel military artificial intelligence. Based on an estimated 30 billion such scans, Niantic Spatial has developed sophisticated geolocation models now being integrated into software by Vantor, a provider of navigation systems for drones and solutions used by various defense entities, including the US Army.
The Genesis of Data Collection: How Pokémon Go Fueled Geospatial AI
Pokémon Go was initially designed to encourage players to explore their physical surroundings. However, recent revelations suggest that the vast datasets collected by the game are now contributing to systems utilized in military contexts.
The AR scanning feature, introduced in Pokémon Go in 2021, incentivized players to film their environments in exchange for additional items and bonuses. While not mandatory, Niantic Spatial sought explicit permission for each recording. To earn these extra rewards, players used their smartphones to scan buildings, streets, and natural features in 360-degree views.
These approximately 30 billion scans served as crucial training data for AI models capable of constructing highly precise, three-dimensional maps of the world. On Niantic’s servers, a colossal database of images, each linked to specific GPS coordinates, began to form. Niantic Spatial then processed these images using AI to create detailed 3D maps that enable navigation even without a GPS signal. This also allowed them to train their own navigation model based on these aggregated scans.
Following the sale of its gaming division, including Pokémon Go, to Scopely for approximately $3.5 billion in 2025, Niantic spun off its geolocation AI segment as Niantic Spatial. This new entity now possesses these highly trained models, powered by historical scans from Pokémon Go and other applications, offering them to business partners across various sectors—from robotics and logistics to defense.
From Mobile Gaming to Drone Navigation Systems
In December 2025, Niantic Spatial announced a significant partnership with Vantor, a company specializing in navigation software for drones and defense systems. The collaboration aims to develop an integrated Visual Positioning System (VPS), designed to enable drones, ground robots, and operators using AR headsets to determine their exact location solely from camera feeds, eliminating reliance on GPS. This capability is vital in environments where satellite signals are jammed, spoofed, or simply unavailable—conditions frequently encountered in modern conflict zones. This development raises important questions about Pentagon AI, military integration, and ethical considerations in data usage.
Technically, the system merges Niantic Spatial’s ground-based infrastructure, which relies on 3D spatial reconstruction, with Vantor Raptor’s aerial software, which analyzes video streams from drones and other airborne platforms. This integration allows diverse sensors—from a drone’s camera to a soldier’s AR glasses—to “perceive” a consistent coordinate system and precisely locate themselves in terrain, even without traditional satellite maps. The potential for such systems to navigate without external signals also touches upon the AI authenticity dilemma, where machine interpretation of the world becomes paramount.
Vantor already holds contracts worth up to hundreds of millions of dollars with the US Army for delivering training software and systems based on high-quality 3D terrain maps, further underscoring the military implications of this project.
Niantic Spatial’s Clarification: Distinguishing Data Sources
Following media reports and public interest, Niantic Spatial issued statements asserting that current data from Pokémon Go is not part of its agreement with Vantor. The company clarified that after the sale of Pokémon Go to Scopely, game data is no longer shared with Niantic Spatial, and the transmission of AR scans ceased as part of the transition process. However, Niantic Spatial did acknowledge that historical scans from Pokémon Go were a contributing factor in training the early versions of the geolocation models now utilized by the company.
Frequently Asked Questions (FAQ)
Niantic Spatial is a company spun off from Niantic (the creator of Pokémon Go) that focuses on geospatial artificial intelligence. It leverages advanced models, initially trained with vast amounts of real-world scan data from games like Pokémon Go, to create detailed 3D maps and enable GPS-independent navigation for various industries, including defense.
Players scanned their environments using an in-game AR feature, accumulating an estimated 30 billion scans. This data was used to train Niantic Spatial’s AI models, enabling the creation of precise 3D maps and navigation systems that do not rely on GPS. These models are now integrated into military-grade drone and navigation software by partners like Vantor.
A Visual Positioning System (VPS) allows devices like drones, robots, and AR headsets to determine their exact location using only camera imagery, without needing GPS. This is critically important for military operations in areas where GPS signals may be jammed, spoofed, or unavailable, providing a robust navigation solution in contested environments.
Niantic Spatial has stated that current Pokémon Go data is not being shared with military partners after the game’s sale to Scopely. However, the company confirmed that historical scans collected from Pokémon Go players were instrumental in training the initial versions of the geospatial AI models it currently uses and offers to partners.
The use of civilian-generated data, even if initially collected for recreational purposes, by military entities raises significant ethical concerns. These include questions about user consent for unexpected applications of their data, privacy implications, the potential for unintended consequences in conflict zones, and the broader societal trust in technology platforms. It highlights the need for transparency and clear policies regarding data ownership and application in the digital age.
Source: The Guardian, Trouw, United24 Media, Drone XL
Opening photo: Krzysztof Wilamowski / Android.com.pl
About Post Author
