Facial Recognition Misidentifies Pregnant Woman Leading to False Arrest in Detroit

Porcha Woodruff was arrested and subsequently had charges dropped due to an unreliable facial recognition match. Despite being visibly pregnant, she was implicated in a robbery and carjacking based on an outdated photo used in a lineup.

Risk classification

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

The facial recognition technology exhibited unequal performance, having significantly higher false positive rates for Black individuals, directly leading to the wrongful arrest of Porcha Woodruff.

Additional risk subdomains

• 7.3 Lack of capability or robustness: The facial recognition system failed to perform reliably, generating an incorrect match from a low-quality video.
• 5.1 Overreliance and unsafe use: Law enforcement officers over-relied on the facial recognition match as the sole basis for a photo lineup and subsequent arrest warrant.

Causal factors

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

The incident was initiated by an incorrect match generated by a deployed facial recognition AI system, which was an unintentional error in its performance.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The report describes the use of facial recognition in law enforcement, which is explicitly classified as a high-risk application under the EU AI Act due to its significant implications for safety, fundamental rights, and public interests.

AI system and alleged parties

• AI system: facial recognition technology
• AI purpose: Face Recognition; Identification
• Behaviour type: Tool
• Alleged developer: unknown
• Alleged deployer: Detroit Police Department
• Alleged harmed parties: Porcha Woodruff

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 Minor
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Minor
• 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 Minor
• Privacy: direct Negligible, indirect Minor
• Psychological: direct Negligible, indirect Minor
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Physical

Reported: The report explicitly describes physical harm to Porcha Woodruff caused indirectly by the incident.

Indirectly caused: Woodruff suffered dehydration and stress-induced contractions requiring hospital treatment after being detained for 11 hours due to the false arrest.

Financial

Reported: The report describes financial impacts, including a personal bond and a settlement in a related case.

Indirectly caused: Woodruff had to secure a 100,000 dollar personal bond and likely incurred legal fees to have the charges dismissed. Robert Williams received a 300,000 dollar settlement from the city.

Inferred additional harm: Woodruff likely incurred significant legal defense costs and medical bills from her hospital visit.

Differential treatment

Reported: The report explicitly describes systematic differential treatment of Black individuals caused directly and indirectly by the incident.

Directly caused: The AI system generated a false positive match for a Black female suspect, reflecting the technology's documented bias and higher error rates for Black individuals.

Indirectly caused: Woodruff, a Black woman, was subjected to wrongful arrest and detention, demonstrating systemic racial bias in the deployment of facial recognition by law enforcement.

Inferred additional harm: Black communities are disproportionately subjected to police surveillance and higher rates of false positive identifications.

Civil rights

Reported: The report explicitly describes violations of human and civil rights caused indirectly by the incident.

Indirectly caused: Woodruff's civil rights were violated through wrongful arrest, 11-hour detention, and prosecution based on an uncorroborated AI match.

Inferred additional harm: The widespread use of facial recognition by police threatens the public's constitutional rights to privacy and freedom of assembly.

Democracy

Reported: The report explicitly describes how facial recognition technology has been used to monitor and identify Black Lives Matter protesters.

Indirectly caused: The use of facial recognition to monitor and identify protesters threatens the democratic right to peaceful protest and public assembly.

Privacy

Reported: The report explicitly describes privacy violations caused indirectly by the incident.

Indirectly caused: Woodruff's personal data (an eight-year-old mugshot) was stored and used in a facial recognition database and lineup without her active consent.

Inferred additional harm: Millions of individuals have had their photos scraped from the internet or stored in police databases, violating their expectation of privacy.

Psychological

Reported: The report explicitly describes psychological distress and trauma caused indirectly by the incident.

Indirectly caused: Woodruff experienced severe emotional distress, anxiety, and trauma from being falsely accused and detained for 11 hours while eight months pregnant.

Inferred additional harm: Her children and fiance likely suffered emotional distress and anxiety from witnessing her arrest.

People affected

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

Potential causes

Management

• Inadequate Officer Training: Management failed to train officers on the limitations of facial recognition.
• Over-reliance on Tech Metrics: Prioritized technology speed over rigorous investigative standards.
• Lack of Oversight Audits: No regular audits of search outcomes to identify misidentifications.

Technology

• Algorithmic Bias in FRT: FRT has poor accuracy rates for Black and female individuals.
• High False Positive Rates: Algorithms produce many false matches, especially on non-white faces.
• Use of Low Quality Images: Grainy surveillance footage reduces accuracy and reliability of matches.

Data Inputs

• Outdated Mugshot Database: System matched suspect using an eight-year-old photo instead of current ID.
• Biased Training Data: Training datasets predominantly feature white faces, causing racial bias.
• Scraped Internet Images: Unregulated scraping of public images leads to unverified database entries.

Human Factors

• Automation Bias: Officers over-relied on the technology's suggestions without verification.
• Inadequate Visual Review: Detectives failed to notice the suspect in video was not pregnant.
• Eyewitness Misidentification: Victim selected Woodruff from a biased photo lineup presented by police.

Process and Methods

• Flawed Photo Lineup Process: Lineups were not double-blind, leading to suggestive identification.
• Lack of Corroborating Evidence: Arrests were made based solely on FRT leads without secondary evidence.
• Poor Investigative Follow-up: Failed to interview other suspects or cross-reference physical traits.

Regulatory Environment

• Absence of Use Limitations: Lack of legal bans or strict limits on police use of biometric surveillance.
• Lack of National Standards: No standardized federal guardrails for deploying facial recognition.
• Unregulated Data Scraping: No laws preventing companies from scraping public photos for databases.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide highly detailed, consistent accounts of Porcha Woodruff's arrest, the role of the facial recognition technology, and the broader context of similar incidents in Detroit. The facts are well-documented through court filings and official statements, leaving little ambiguity about the sequence of events or the technology's failure.
• Ambiguities identified: The specific facial recognition software brand used in Woodruff's case is not explicitly confirmed, though general vendors are discussed.
• Alternative interpretations: One could argue the primary failure was human investigative error rather than AI failure, though the AI's false match was the indispensable trigger for the chain of events.

A pregnant woman was falsely arrested in Detroit due to an incorrect facial recognition match. While representing a serious violation of civil rights and highlighting systemic algorithmic bias in municipal policing, the incident remains a localized domestic law enforcement issue with minor national security implications.

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

Threat characteristics

• Imminence: Long-term. The misuse of facial recognition by law enforcement represents an ongoing domestic policy and civil liberties concern rather than an immediate national security crisis.
• Autonomy: Human-supervised. The AI system functioned as a search tool, but human police officers and detectives made the final decision to conduct the lineup and execute the arrest.
• Novelty: Established threat. This is the third known false arrest in Detroit and sixth in the United States arising from facial recognition mismatches, representing an established failure mode.

Impact by dimension

• Physical security: Negligible. No physical infrastructure, kinetic weapons, or life-safety control systems were targeted or compromised in this incident.
• Information security: Negligible. The incident did not involve classified systems, state-sponsored information operations, or the compromise of national intelligence methods.
• Sovereignty: Negligible. While involving a local government agency (municipal police), this incident does not threaten state authority, territorial control, or core federal government functions.
• Economic security: Negligible. The incident does not impact national economic stability, critical supply chains, or strategic technological competitive advantage.
• Societal stability: Minor. The incident involves a serious civil rights violation and highlights algorithmic bias against Black individuals, but it is a localized domestic policing issue manageable through legal and policy reforms rather than a threat to national stability.