A state that loses the capacity to screen at its own gates does not contain the consequence within its own borders. Canada’s immigration bureaucracy is now creating a security problem not only for Canada, but for the United States.

For two centuries, Washington has benefited from an assumption it rarely had to inspect: that the Canadian state could competently govern its own gate. The northern border was not treated as a major strategic vulnerability because the two countries operated on a quiet premise. Whoever Canada admitted, regularized, naturalized, or failed to remove was generally not someone the United States needed to screen again as a primary threat.

That premise is now eroding. The arrangement built upon it has not yet been repriced.

The issue is not immigration as such. Lawful immigration strengthens North America. Genuine refugees are not the problem; in fact, they are harmed by a system that cannot distinguish them from hostile actors. Nor is this an argument against any religious or ethnic community. The issue is narrower, more serious, and more defensible: whether Canada’s immigration system can identify and act against individuals connected to terrorism, hostile foreign influence, serious criminality, or organized ideological violence before status is granted or retained.

A gate that cannot distinguish is not merely inefficient. It is dangerous.

The conventional complaint against Immigration, Refugees and Citizenship Canada is that it is slow. That is the wrong metric. Speed measures throughput. Security depends on discrimination in the literal sense: the state’s ability to tell one case from another and to weight them differently. A fast gate that cannot distinguish is more dangerous than a slow one.

The real failure appears at the seam where status changes: applicant to resident, resident to citizen, inadmissible person to removed person. At that point, the state is supposed to perform its most basic sovereign act: decide who may remain inside the political community and who may not. If available information exists but does not reach the decision point, the gate is not functioning.

Recent cases show why this matters beyond Canada.

In May 2024, Canada granted citizenship to Ahmed Eldidi. In July 2024, Canadian authorities arrested him and his son in connection with an alleged mass-casualty plot. Public reporting states that the father was alleged to have appeared in a 2015 Islamic State propaganda video, and that the video was not available to the officials who adjudicated the file. The allegations remain unproven in court. But the institutional issue is separate: if disqualifying information existed and did not reach the citizenship decision point, intelligence and immigration adjudication were not operating as one security system.

A second case shows how Canadian failure can transfer risk directly toward the United States. In September 2024, a Pakistani national admitted to Canada on a study permit was arrested near Ormstown, Quebec, close to the New York border. According to the United States Department of Justice, his objective was to cross into the United States and conduct a mass shooting at a Jewish center in New York City. The strategic significance is exact: Canadian admission, American target, northern-border approach.

A third failure concerns removal. Canada listed Iran’s Islamic Revolutionary Guard Corps as a terrorist entity in June 2024, years after the United States did so and after the destruction of Flight PS752 killed dozens of Canadian citizens and permanent residents. Designation is not the same as removal. A state can correctly identify a threat category and still fail to act against individuals linked to it. Public reporting and parliamentary material indicate that, as of early 2026, removals of regime-linked inadmissible individuals remained extremely limited. If designation does not translate into removal, it becomes documentation of incapacity rather than an instrument of security.

These are not identical errors. They are different failure modes: information not integrated before status is granted, Canadian territory used as a staging area for a U.S.-directed threat, and legal designation not followed by meaningful enforcement. Together, they point to a deeper problem. Canada is losing the reliable ability to convert known or knowable security information into immigration decisions.

That is not merely Canada’s problem. Canada is America’s northern perimeter.

The United States and Canada are functionally integrated in ways the public debate rarely acknowledges. They share intelligence architecture, air-defense responsibilities through NORAD, law-enforcement coordination, trade corridors, infrastructure, and geography. Each layer assumes partner competence. When Canadian screening degrades, American exposure rises.

The risk is not limited to physical border crossing. Networks operating from Canadian territory can finance, recruit, propagandize, intimidate diaspora communities, or support hostile activity without crossing the border themselves. But the border route also exists. The Ormstown case shows that Canadian territory can become the final approach to an American target.

A serious U.S. posture should therefore begin with a simple adjustment: stop pricing the Canadian gate as automatically sound.

This does not require hostility toward Canadians. It requires seriousness about state capacity. Washington should formally assess whether Canadian immigration-screening failures create northern-border homeland-security exposure. The two governments should audit whether immigration adjudicators receive available intelligence, law-enforcement, and public-safety information before permanent status or citizenship is granted. U.S. authorities should provide clearer public reporting on northern-border terrorism-watchlist encounters and trends. Canada should not grant long-term status where unresolved terrorism, hostile-state, or serious-criminal indicators remain outstanding. And when Canada designates an entity as terrorist, it should report, within lawful limits, how many linked inadmissible persons are identified, removed, pending removal, or still present.

None of this is anti-immigration. It is pro-sovereignty. A serious immigration system must be capable of distinction. If it cannot distinguish lawful entrant from threat, refugee from radical, citizen from hostile actor, and designation from enforcement, then it has ceased to govern the gate.

The United States did not create Canada’s screening failure and cannot fully correct it from its own side of the border. But it can stop pretending that Canadian gatekeeping is merely a Canadian domestic issue. It can treat the northern perimeter as a security system whose reliability must be verified, not assumed.

States resolve exposures in one of two ways. They reprice them deliberately through reform while the choice remains available, or they are repriced by an event that removes the choice. Canada’s record suggests a preference for delay. Washington should not share it.

(aka Seraphima) is the founder of the Seraphima Institute for North American Coalition Strategy (SINACS), a U.S.-owned think tank focused on continental security, Canada-U.S. realignment, border integrity, and threats to Western institutions. She is a Bangladeshi-born writer and researcher based in Canada, and a former Queen’s University engineering student.

This article was originally published by RealClearDefense and made available via RealClearWire.