Full Report
An estimated 100 million people live with facial differences. As face recognition tech becomes widespread, some say they’re getting blocked from accessing essential systems and services.
Analysis Summary
# Main Topic
Widespread deployment of facial recognition and identity verification technologies is resulting in systemic failure to accurately identify individuals living with facial differences, leading to denied access to essential services and increased stigma.
## Key Points
- An estimated 100 million people globally with facial differences are being systematically excluded by identity verification systems powered by machine learning.
- Specific examples include being blocked from updating driver's licenses (DMV access), accessing financial services, and using essential tools like phone face-unlock systems and social media filters.
- The issue stems from facial recognition algorithms, which are often trained on datasets that do not adequately represent diverse facial structures resulting from congenital conditions or injuries.
- The technology failure echoes real-world stigma, with victims stating the systems tell them "that I don’t have a human face."
- While police face recognition systems are often noted for bias against Asian and Black people, this specific technological failure affects those with craniofacial conditions (e.g., Freeman-Sheldon syndrome).
## Threat Actors
- Not applicable. The issue is systemic failure within commercial and governmental identity verification technology rather than malicious threat actor activity.
## TTPs
- **Deployment of Inaccurate ML/AI Authentication:** Use of various face checking methods (selfie comparison to ID, liveness tests) which fail to properly authenticate users with non-typical facial features.
- **Systemic Exclusion:** The failure of multiple access points (government services, financial systems, consumer tech) creates an environment of exclusion for those affected.
## Affected Systems
- Department of Motor Vehicles (DMV) identification and photo update systems.
- Financial services identity verification systems.
- Consumer technologies: Social media filters, phone face-unlock systems.
- Airport passport gates.
- Video conferencing software (background blurring features).
## Mitigations
- **Advocacy and Visibility:** Organizations like Face Equality International (FEI) are campaigning for recognition of facial equality as a human right.
- **System Auditing/Improvement:** Need for development teams to address failures in facial recognition algorithms to ensure inclusivity for the facial difference community.
- **Alternative Verification:** Systems must implement robust, non-biometric fallback options for service access where facial recognition fails (as highlighted by the DMV experience).
## Conclusion
The rapid proliferation of facial recognition creates a significant digital barrier for an estimated 100 million people with facial differences, translating technical inaccuracy directly into social and civic exclusion. Organizations implementing these technologies must prioritize inclusive training data and provide immediate, accessible non-biometric alternatives to resume essential functionality for affected individuals.