.avif){kind=small}
The Pakistani National Database and Registration Authority (NADRA) implemented a biometric identification system that, while intended to improve service delivery and security, has caused widespread harm. Automated rules within the system have led to the mass suspension of hundreds of thousands of national identity cards (CNICs), effectively rendering citizens stateless and denying them access to basic rights, welfare, and financial services. Courts have repeatedly ruled that NADRA's automated adjudication of family and paternity disputes is illegal and lacks jurisdiction.
When the leader of the Afghan Taliban was found possessing a valid ID card in the Pakistani national biometric identification database system, Pakistan launch a national re-verification campaign that is linked to numerous changes in recognition status and loss of services.
The system's rigid automated rules and biometric matching failures (such as facial recognition errors and cascading family tree blocks) demonstrate a lack of capability to robustly handle real-world complexities.
The widespread harms resulted from the post-deployment execution of automated database rules and biometric matching algorithms, which produced unexpected negative outcomes.
Risk Level 2: High Risk. The system involves biometric identification and facial recognition used by government authorities, which has significant implications for fundamental rights and access to essential public services.
Severity ratings are based on this scale.
Highest direct severity in any category: Severe. 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.
Reported: Yes, the report describes an incident where a minor was raped by an employee who obtained her phone number from the database.
Indirectly caused: A 15-year-old girl was raped by a welfare employee who used the phone number retrieved from her records in the database to locate and harass her.
Reported: Yes, the report describes court-ordered fines and individual financial losses due to blocked cards.
Directly caused: The Islamabad High Court fined NADRA Rs 500,000 (approximately $2,886) to be paid to Urooj Tabani for illegally blocking her card.
Indirectly caused: Gulzar Bibi and other affected citizens lost daily wages and spent scarce resources traveling to resolve their blocked cards, while also losing access to welfare funds.
Inferred additional harm: Across 660,000 blocked individuals, the cumulative financial loss from lost wages, travel costs, and denied welfare benefits likely amounts to millions of dollars.
Reported: Yes, the report describes differential treatment of manual laborers and ethnic minorities.
Directly caused: Manual laborers with worn fingerprints faced authentication failures because the biometric readers could not scan their prints.
Indirectly caused: The Pashtun community was disproportionately affected, accounting for 63% of blocked CNICs in 2017 despite being only 15% of the population.
Inferred additional harm: Other marginalized groups, such as the Bengali community and transgender individuals, likely faced systemic barriers and differential treatment due to rigid database rules.
Reported: Yes, the report describes widespread violations of fundamental human and civil rights.
Directly caused: The automated blocking of CNICs directly stripped citizens of their legal identity, rendering them effectively stateless and denying them rights to vote, travel, and access education.
Indirectly caused: The system was used punitively to block the CNICs of civil rights leaders like Manzoor Pashteen and Mohsin Dawar, suppressing their freedom of movement and expression.
Inferred additional harm: Hundreds of thousands of citizens had their fundamental civil rights suspended without due process due to automated database flags.
Reported: Yes, the report describes the erosion of democratic norms through political targeting and minority tracking.
Indirectly caused: The system eroded democratic norms when courts and state agencies used it to block the CNICs of opposition politicians and civil rights leaders, and to track religious minorities (Ahmadis) to prevent them from holding public office.
Inferred additional harm: The existence of a centralized, unregulated biometric database creates a powerful tool for state surveillance and political suppression, undermining democratic processes at scale.
Reported: Yes, the report describes massive data leaks and unauthorized access to personal information.
Indirectly caused: The names, addresses, and CNIC numbers of over 100 million Pakistanis were leaked and sold online, and employees abused their access to retrieve private phone numbers.
Inferred additional harm: The lack of a robust data protection law in Pakistan means that the biometric and personal data of nearly the entire population remains highly vulnerable to further leaks and unauthorized surveillance.
Reported: Yes, the report explicitly describes severe emotional distress, trauma, and anxiety caused by the system's actions.
Directly caused: Urooj Tabani suffered unimaginable pain, agony, emotional distress, and psychological trauma after the system blocked her CNIC and disputed her paternity.
Indirectly caused: Gulzar Bibi and her family experienced a permanent sense of dread and anxiety due to their identities being suspended and facing potential statelessness.
Inferred additional harm: Given that nearly 660,000 individuals had their CNICs blocked, it is highly likely that tens of thousands of others suffered severe psychological distress and anxiety.
The automated blockings and data leaks within Pakistans biometric database system (NADRA) led to severe civil rights violations, rendering over 600,000 citizens stateless and exposing personal data of 100 million people. This highlights the severe national security and societal stability risks of unregulated state-level algorithmic identity management.