October 30, 2025
Crossing the Line? The Canada Border Services Agency’s Traveller Compliance Indicator and the Lesson from New Zealand
When the Canada Border Services Agency (CBSA) introduced its Traveller Compliance Indicator (TCI) in September 2025, it was hailed as a step toward smarter, data-driven border management. The AI-powered system predicts how likely travellers are to comply with Canadian law, using years of historical data and behavioral patterns. Yet, as Canada embraces predictive analytics at the border, New Zealand’s experience offers a cautionary tale. Its early use of algorithmic “risk scoring” in immigration sparked public backlash, policy reform, and eventually the creation of an Algorithm Charter a framework that set global standards for fairness, transparency, and accountability in government use of AI. The question for Canada is whether it will learn from those lessons before the line between innovation and intrusion is crossed.[1]
How the TCI Works
The TCI draws on up to five years of CBSA records, incorporating data such as travel frequency, type of identification, vehicle details, and prior entry patterns. Currently in pilot testing at selected land ports of entry, the CBSA plans to expand the system nationwide in the coming years. While the agency promotes TCI as a means of enhancing border efficiency and national security, its predictive nature raises significant legal, ethical, and privacy concerns. [2]
Lessons from New Zealand
New Zealand faced similar challenges several years earlier. After developing predictive “harm models” that analyzed variables such as age, gender, ethnicity, immigration status, and use of public services, the government encountered widespread criticism for bias and lack of transparency.[3] The public backlash led to the creation of the Algorithm Charter for Aotearoa New Zealand (2020), which set clear principles for fairness, accountability, and human oversight in government use of algorithms.[4]
The Canadian experience with the TCI bears striking resemblance to New Zealand’s earlier trials. The risk is that predictive models when based on sensitive or poorly contextualized data can reproduce existing inequities, disproportionately flagging racialized or marginalized travellers as “high-risk.”
Potential Implications for Canada
While the TCI promises smoother border management, it simultaneously opens the door to algorithmic bias, false positives, and opaque data practices. Predictive modelling without informed consent, cross-system data sharing, and the aggregation of personal histories all raise profound privacy and human rights concerns.[5] These issues demand rigorous, plain-language privacy and algorithmic impact assessments before full deployment.
Conclusion
If left unchecked, the TCI’s pursuit of “non-compliant” travellers could end up profiling the very people it seeks to protect creating new risks for Canada’s border integrity and public trust. Canada should follow New Zealand’s lead and establish clear, transparent guardrails on the use of AI in public governance to ensure that security never outweighs fairness or accountability.
