Is Pokémon Go racist? How the app may be redlining communities of color

Through a crowdsourcing social media campaign in 2016, several journalists and researchers demonstrated that augmented reality locations in the popular smartphone game Pokemon Go were more likely to be in white neighborhoods.

Risk classification

• Primary risk domain: 1 Discrimination & Toxicity
• Primary risk subdomain: 1.3 Unequal performance across groups

The game's mapping system performed unequally across demographic groups, providing significantly fewer PokéStops in minority neighborhoods due to biased historical training data.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The unequal distribution of PokéStops was caused by the automated mapping system's reliance on biased historical data, which was an unintended outcome of Niantic's goal to deploy the game quickly.

EU AI Act risk tier

• Risk tier: 4 Minimal or No Risk

Minimal or No Risk: The system is an entertainment application (mobile game), which poses low or negligible risks to users and society under the EU AI Act framework.

AI system and alleged parties

• AI system: Pokémon Go (Niantic)
• AI purpose: Resource Allocation; Game Content Generation
• Behaviour type: Tool
• Alleged developer: Niantic Labs
• Alleged deployer: Niantic Labs
• Alleged harmed parties: non-white neighborhoods, communities of color

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Minor, indirect Minor
• Civil rights: direct Negligible, indirect Minor
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Minor
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Differential treatment

Reported: The report explicitly describes systematic differential treatment of minority communities who had significantly fewer PokéStops in their neighborhoods.

Directly caused: The Pokémon Go mapping system generated an average of 55 PokéStops in majority-white neighborhoods compared to only 19 in majority-Black neighborhoods, directly resulting in unequal access to game resources.

Indirectly caused: Black players faced disproportionate suspicion and questioning by police while playing the game in public spaces, as illustrated by a police stop of two Black men playing in a park.

Inferred additional harm: We infer that millions of minority residents in the US experienced systemic exclusion and unequal access to the game's features and economic benefits due to their geographic location.

Civil rights

Reported: The report explicitly describes concerns regarding racial profiling and unequal access to services, which touch on civil rights.

Indirectly caused: The game's mechanics indirectly exposed Black players to racial profiling and unwarranted police stops, violating their civil right to equal treatment and freedom from arbitrary detention.

Inferred additional harm: We infer that multiple Black players experienced unwarranted police stops or questioning while playing, representing localized civil rights infringements.

Psychological

Reported: The report explicitly describes psychological distress and anxiety experienced by Black players due to systemic racial profiling while playing.

Indirectly caused: Black player Omari Akil reported experiencing severe anxiety and distraction during a 20-minute walk, spending 14 minutes worrying about being perceived as suspicious by police or residents while playing.

Inferred additional harm: We infer that thousands of Black players experienced similar anxiety, hypervigilance, and distress when playing the game in public spaces due to systemic racial profiling.

People affected

• Occurrences reported: 1
• People reportedly harmed: 4
• People reportedly exposed: 50000000

Potential causes

Management

• Prioritizing Global Launch: Management prioritized rapid deployment over addressing geographic bias.
• Delayed Corrective Action: Niantic deferred addressing PokeStop disparities until after global launch.

Technology

• Legacy Map Reusage: Niantic reused map data from Ingress to populate PokeStops.
• Geospatial Data Dependence: The game relied heavily on historical geospatial coordinates.

Data Inputs

• Biased Crowdsourced Inputs: Map data was sourced from Ingress players who were mostly white and male.
• Demographic Imbalance: Lack of diverse user inputs led to fewer stops in minority areas.

Human Factors

• Racial Profiling Risks: Black players face suspicion and police stops while playing in public.
• Homogeneous User Base: Ingress player demographics shaped the original geographic map layout.

Process and Methods

• Lack of Map Bias Review: Niantic did not evaluate the racial equity of reused geospatial data.
• Frozen PokeStop Requests: Niantic suspended new PokeStop requests during the initial launch phase.

Regulatory Environment

• No AR Equity Regulations: No regulatory standards govern fair distribution of digital resources.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent accounts of the digital redlining issue, backed by empirical data from the Urban Institute (55 vs 19 PokéStops). The technical root cause (reliance on Ingress crowdsourced data) is clearly explained by the developer and independent researchers. The safety concerns for Black players are well-documented through first-hand essays and anecdotes.
• Ambiguities identified: The exact number of players who had to pay to play or who faced police questioning is not fully quantified, relying instead on qualitative anecdotes and statistical averages of PokéStop distribution.
• Alternative interpretations: The disparity could be interpreted purely as a legacy data issue rather than an active algorithmic bias, though the resulting impact remains discriminatory.

An augmented reality game's map layout exhibited digital redlining due to reliance on biased historical crowdsourced data. While raising valid civil rights concerns regarding resource distribution and potential racial profiling during public play, the incident has negligible national security implications.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents ongoing strategic concern regarding algorithmic bias in consumer technologies rather than an active crisis.
• Autonomy: Human-controlled. The system maps virtual assets based on historical parameters without autonomous agency or physical action.
• Novelty: Established threat. Algorithmic bias stemming from unrepresentative historical training data is a well-documented and established threat.

Impact by dimension

• Physical security: Negligible. No impact on physical security, kinetic weapons, or critical infrastructure.
• Information security: Negligible. No evidence of information warfare, state-sponsored intelligence compromise, or systematic information manipulation.
• Sovereignty: Negligible. No threat to state sovereignty, territorial control, or core government decision-making processes.
• Economic security: Negligible. No strategic economic or technological threat to national industries, financial systems, or competitive advantage.
• Societal stability: Minor. Systemic bias in geospatial resource distribution led to localized civil rights concerns and potential racial profiling risks for minority players.