CBSA launches ReportIn app with facial recognition for deportee monitoring
The Canada Border Services Agency launched the ReportIn app in 2024 to allow people ordered to leave Canada to report remotely using facial recognition. The system raises concerns about algorithmic transparency and bias, though CBSA reports testing shows 99.9 percent accuracy across demographic groups.
What happened
The Canada Border Services Agency (CBSA) launched the ReportIn app in fall 2024 to allow foreign nationals and permanent residents subject to immigration enforcement conditions to report without visiting a CBSA office in person. The app uses facial biometrics to confirm identity and records location data. The solution has been in development since 2021, replacing voice biometric technology that had been phased out due to failures.
The app relies on Amazon Web Services and facial recognition technology. Although the app is voluntary, CBSA has kept the facial recognition algorithm as a trade secret, and the specific facial recognition vendor has not been publicly named.
Reported impact
The CBSA's use of facial biometrics raises questions about user consent, algorithmic bias, and decision-making transparency. Critics note that keeping the algorithm as a trade secret may leave people without the right to understand how decisions are made about them.
Index analysis
According to an algorithmic impact assessment, Credo AI tested the facial recognition software for bias and found a 99.9 percent facial match rate across six different demographic groups. CBSA states the app will be continuously tested after launch to assess accuracy and performance, with final decisions made by humans.
Amazon Rekognition has also been tested by NIST-accredited iBeta, with Rekognition Face Liveness confirmed compliant to the ISO/IEC biometric presentation attack detection standard.
Governance controls
CBSA indicates that individuals will not be tracked constantly; their location will be collected on a limited basis. Final enforcement decisions will be made by humans rather than algorithms alone.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.