Target Suggested Maternity-Related Advertisements to a Teenage Girl's Home, Allegedly Correctly Predicting Her Pregnancy via Algorithm

Target recommended maternity-related items to a family in Atlanta via ads, allegedly predicting their teenage daughter’s pregnancy before her father did, although critics have called into question the predictability of the algorithm and the authenticity of its claims.

Risk classification

• Primary risk domain: 2 Privacy & Security
• Primary risk subdomain: 2.1 Compromise of privacy by obtaining, leaking or correctly inferring sensitive information

The AI system compromised user privacy by correctly inferring highly sensitive, non-public medical information (pregnancy status) from routine shopping data without explicit consent.

Additional risk subdomains

• 4.3 Fraud, scams, and targeted manipulation: Target deliberately manipulated the presentation of coupons by mixing in random items to obfuscate their tracking and influence consumer behavior without their knowledge.

Causal factors

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

The privacy risks and behavioral profiling were caused by the post-deployment operation of Target's predictive analytics algorithm, which was intentionally designed to infer pregnancy status.

EU AI Act risk tier

• Risk tier: 4 Minimal or No Risk

Minimal or No Risk: The system is a commercial marketing and predictive analytics tool used for product recommendation and targeted advertising, which does not fall under the prohibited or high-risk categories of the EU AI Act.

AI system and alleged parties

• AI system: Target pregnancy prediction algorithm
• AI purpose: Marketing Analytics Generation; Product Recommendation
• Behaviour type: Assistant
• Alleged developer: Target
• Alleged deployer: Target
• Alleged harmed parties: Target customers

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 Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Differential treatment

Reported: The report describes Target profiling and targeting a specific demographic group (pregnant women) for marketing purposes.

Directly caused: Target assigned pregnancy prediction scores to female shoppers and sent them targeted coupon booklets based on this classification.

Privacy

Reported: The report explicitly describes privacy concerns and violations resulting from Target's predictive modeling.

Directly caused: The AI system inferred a high school girl's pregnancy status from her shopping data and sent targeted maternity coupons to her home, revealing her private medical condition to her father.

Inferred additional harm: We infer that thousands of other women had their private, non-public pregnancy status correctly inferred and tracked by Target's database without their explicit consent.

People affected

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

Potential causes

Management

• Growth Prioritized Over Privacy: Management pushed to capture the lucrative mom and baby segment early.
• Inadequate Risk Assessment: Failed to foresee the public relations and ethical risks of creepy ads.
• Lack of Corporate Transparency: Management hid predictive practices and blacklisted inquiring reporters.

Technology

• Pregnancy Prediction Model: Algorithm analyzed 25 products to predict pregnancy and due dates.
• Intrusive Behavioral Profiling: Model inferred highly sensitive personal health states from purchase habits.
• Lack of Algorithmic Transparency: Model operated without clear validation metrics shared with the public.

Data Inputs

• Guest ID Tracking System: Linked credit cards, emails, and purchase histories to track habits.
• Purchasing Third-Party Data: Target bought external demographic and personal data to profile guests.
• Unconsented Health Data Mining: Used routine retail purchases to infer clinical-level health changes.

Human Factors

• Customer Creepiness and Discomfort: Customers felt violated when Target knew of their pregnancies in advance.
• Family Conflict from Disclosure: Targeted coupons revealed a teenager's pregnancy to her father.
• Skepticism of Overreaching AI: Public backlash against perceived corporate spying and predictive profiling.

Process and Methods

• Uncamouflaged Direct Marketing: Initially sent direct baby ads without mixing in random products.
• Lack of Opt-In for Profiling: No process for customers to consent to sensitive health state prediction.
• Defensive Information Siloing: Target cut off communication and banned journalists to hide practices.

Regulatory Environment

• Gaps in Privacy Regulations: Laws did not restrict commercial profiling of non-clinical health data.
• Permissive Data Broker Market: Lack of regulation allowed Target to buy extensive personal demographic data.
• Weak Enforcement on Profiling: No regulatory oversight prevented predictive modeling of minors.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide detailed, first-hand accounts from the NYT journalist who interviewed the statistician Andrew Pole and Target employees. While the specific anecdote about the father is noted as potentially apocryphal or exaggerated by some commentators, the technical details of the predictive model, the data points used, and Target's corporate response are well-documented.
• Ambiguities identified: The exact veracity of the high school girl anecdote is debated, and the precise mathematical formulation of the 25-product pregnancy prediction score is proprietary and not fully disclosed.
• Alternative interpretations: The incident could be interpreted purely as a corporate data privacy issue rather than an AI safety failure, given that the technology used was traditional predictive statistics rather than modern deep learning.

Target's use of predictive analytics to infer customer pregnancies represents a landmark early case of corporate data profiling invading personal privacy. While it raised significant ethical and societal questions about algorithmic surveillance, its direct national security impact remains minor and manageable under standard consumer privacy and regulatory frameworks.

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

Threat characteristics

• Imminence: Long-term. The incident represents a historical, ongoing strategic concern regarding corporate data privacy rather than an active, immediate crisis.
• Autonomy: Human-supervised. The predictive analytics model generated the pregnancy scores autonomously, but human marketers designed the advertising campaigns and modified the targeting strategy.
• Novelty: First-of-its-kind. This was one of the earliest and most prominent public examples of a corporation using predictive algorithms to infer highly sensitive, non-public personal health data.

Impact by dimension

• Physical security: Negligible. The incident involved commercial marketing data and did not target or impact physical systems, critical infrastructure, or human safety.
• Information security: Negligible. There is no indication of intelligence compromise, state-sponsored information warfare, or classified data theft.
• Sovereignty: Negligible. The system was operated by a private retailer for commercial purposes and did not impact government operations or state sovereignty.
• Economic security: Negligible. While Target experienced revenue growth, the incident did not threaten national economic stability, strategic industries, or critical supply chains.
• Societal stability: Minor. The incident highlights early corporate privacy risks and mass behavioral profiling of sensitive health-related events, though it remains manageable within standard consumer protection frameworks.