Identical Twins Can Open Apple FaceID Protected Devices

Apple's iPhone FaceID can be opened by an identical twin of the person who has registered their face to unlock the phone.

Risk classification

• Primary risk domain: 7 AI system safety, failures, & limitations
• Primary risk subdomain: 7.3 Lack of capability or robustness

The incident stems from a technical limitation where the facial recognition system lacks the capability and robustness to reliably distinguish between identical twins under normal operating conditions.

Additional risk subdomains

• 2.1 Compromise of privacy by obtaining, leaking or correctly inferring sensitive information: The lack of robustness in Face ID directly results in unauthorized individuals gaining full access to the owner's private device and personal data.

Causal factors

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

The security bypass was caused by the Face ID AI system failing to correctly differentiate identical twins, which was an unintentional and unexpected outcome of its deployment.

EU AI Act risk tier

• Risk tier: 4 Minimal or No Risk

Minimal or No Risk: The system is a consumer biometric authentication feature used for unlocking personal smartphones, which poses minimal risk to society.

AI system and alleged parties

• AI system: Face ID (Apple)
• AI purpose: Face Recognition; Identification
• Behaviour type: Tool
• Alleged developer: Apple
• Alleged deployer: Apple
• Alleged harmed parties: People with Twins

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 Negligible, 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

Privacy

Reported: The reports explicitly describe privacy concerns and unauthorized access to personal data due to Face ID failing to distinguish between twins, siblings, or colleagues.

Directly caused: In at least three real-world cases (the Russian twins, the Chinese colleague, and the US mother's 10-year-old son), unauthorized individuals were able to unlock the owner's iPhone X and access their private data.

Inferred additional harm: It is highly likely that many other identical twins or close relatives globally experienced unauthorized access to their devices, compromising their personal data privacy.

People affected

• Occurrences reported: 8
• People reportedly harmed: 4
• People reportedly exposed: 19

Potential causes

Management

• Prioritizing Speed Over Safety: Replacing Touch ID entirely before resolving twin and child security issues.
• Inadequate User Warnings: Failing to clearly warn consumers in user manuals about twin vulnerabilities.

Technology

• Twin Feature Similarity: Similar facial structures of identical twins confuse the 3D mapping sensor.
• Undeveloped Child Features: Children under 13 have less distinct facial features, raising error rates.
• Spoofing via 3D Masks: The neural network can be fooled by low-cost 3D printed masks.

Data Inputs

• Passcode Training Vulnerability: Failed matches followed by passcodes accidentally train the model on twins.
• Single Point Enrollment: Initial setup relies on a single enrollment scan, limiting data diversity.
• Infrared Blockage: Facial coverings that block infrared light prevent accurate data capture.

Human Factors

• Sibling Passcode Sharing: Siblings entering passcodes on failed attempts retrains the AI on their face.
• User Overreliance on Face ID: Users assume absolute security, ignoring risks of lookalikes unlocking phones.
• Intentional System Defeat: Twins or close relatives actively trying to bypass the biometric lock.

Process and Methods

• Inadequate Twin Testing: Insufficient validation against identical twins during algorithm development.
• Software Update Regression: Updates like iOS 12.1 introduced bugs making Face ID temporarily unavailable.
• Lack of Multi-Modal Backup: System relies solely on facial features without backup biometric options.

Regulatory Environment

• No Biometric Warnings Required: Lack of regulatory pressure to warn users about twin-related security gaps.
• Lack of Biometric Regulations: Absence of legal standards for facial recognition security on consumer tech.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of multiple independent tests and real-world incidents confirming that Face ID can be bypassed by identical twins and close relatives, which is also officially acknowledged by Apple.
• Ambiguities identified: None of significance.
• Alternative interpretations: None.

Apple's Face ID launch on the iPhone X revealed biometric vulnerabilities where identical twins and close relatives could bypass authentication. While raising localized privacy concerns for specific user demographics, the incident has negligible national security implications and can be managed through standard security settings like passcodes.

• Overall national security impact: Minor
• Scope: Multiple nations
• Primary target: No clear primary
• Other affected: Multiple nations
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing biometric limitation and privacy concern rather than an active crisis requiring urgent national response.
• Autonomy: Human-controlled. Face ID operates as a biometric tool assisting the user, who retains ultimate control over device settings and authentication choices.
• Novelty: Evolved capability. Represents an advancement of existing biometric bypass threats, shifting from fingerprint spoofing to 3D facial recognition evasion.

Impact by dimension

• Physical security: Negligible. No impact on physical security, critical infrastructure, or kinetic systems.
• Information security: Negligible. No evidence of intelligence compromise, classified model theft, or systematic disinformation campaigns.
• Sovereignty: Negligible. No impact on state authority, government decision-making, or electoral systems.
• Economic security: Negligible. Limited to minor consumer disputes and individual device refunds, with no systemic threat to financial systems or strategic technology.
• Societal stability: Minor. Minor privacy vulnerability affecting specific demographic groups like twins and minors, manageable by reverting to standard passcode security.