Immigration officials are experimenting with an AI tool that advises newcomers on the best place to settle in Canada, one of many ways it’s increasingly being used. Artificial intelligence, including fraud detection.
Stanford University’s Geomatch algorithm is designed to predict the probability of success for new immigrants, including refugees, at a destination country location. It uses machine learning and includes information about past immigrants and their experiences, as well as new arrivals’ work history, education, and personal characteristics.
the use of algorithms by Immigration, Refugees and Citizenship Canada; This was unveiled in the federal department’s new first-ever AI strategy, in partnership with Stanford University’s Migration Policy Institute. The U.S. university declined to comment.
The strategy states that migrants are not obliged to follow the advice, but it is “provided as useful information for migrants to consider”.
Immigration Bureau suspends visa program for foreign entrepreneurs due to abuse
IRCC is experimenting with a number of AI tools, many of which are focused on fraud prevention. The strategy document states that artificial intelligence is helping the department detect false narratives. These could, for example, form the basis for asylum claims or divert attention from security concerns.
Machine learning tools are also being used to detect anomalies in applications and irregular travel patterns, which could indicate that a refugee or immigrant is from a country other than the one they claim.
AI system They are trained to detect the manipulation of documents such as academic records and bank statements, as well as artificially “distorted” photos that can be used to commit identity fraud or mislead immigration officials about one’s age.
“IRCC uses a variety of tools, including advanced analytics, to assist police officers in identifying potential wrongdoing,” department spokeswoman Isabelle Dubois said in an email.
Last year, IRCC investigated an average of about 8,000 suspected immigration fraud cases a month and rejected an average of 7,900 fraudulent applications a month, Dubois said. The proportion of tourist visas refused due to fraud rose from 4.6% in 2024 to 7.1% in 2025.
2025 federal budget: Immigration minister says foreign nationals living in Canada will be prioritized for permanent residency
Under the new strategy, the Immigration Bureau does not aim to implement “fully autonomous AI”; “We are ready to accelerate the adoption of AI systems, which can improve the efficiency of daily tasks that consume significant time and resources.”
Artificial intelligence is already being used to “triage” and evaluate immigration applications, For example, a recommendation as to whether an immigration officer should approve a visitor’s visa.
The ministry says it will continue to use AI to speed up decision-making. For example, by flagging “simple low-risk files” by authorities.
However, the strategy says IRCC does not use autonomous AI agents that can deny client applications. And the department is avoiding “black box” AI models, where the inner workings of the system are mysterious and the logic used is opaque or unexplainable.
“Doing so would violate the principle of administrative law that clients are entitled to a meaningful explanation of their decisions and a transparent appeals process,” the report said.
The strategy calls IRCC a “leader in AI experimentation” at the forefront of integrating artificial intelligence into government operations. He also acknowledged that while AI offers immense potential, the technology also comes with risks, including perpetuating discrimination.
According to the strategy, the department strives to train AI systems using diverse data to prevent potential bias. We also recognize the importance of privacy management, such as having AI systems process only the minimum amount of personal information necessary. Perform regular testing and auditing.
CD Howe Institute report says asylum decisions handed down without hearing raise security and fraud concerns
When developing AI systems, IRCC will integrate privacy measures, including “attempting to use anonymized or internally generated synthetic data everywhere.” If this is not possible, we will still allow you to delete or correct personal information entered into our AI systems.
Saeed Hassan, executive director of the Migrant Workers Alliance for Change, urged caution in using AI as it often makes mistakes.
He said it was “strange” that the ministry was using AI tools to suggest where people should settle rather than relying on advice from the community, as immigration applications do not include detailed information about a person’s likes and dislikes, such as whether they are interested in the arts.
He warned that where AI is involved in analysis, authorities may be inclined to scrutinize files less closely or check facts before making decisions, and IRCC needs to ensure checks and balances are in place.
“People are entrusting their thinking to AI,” he says.
Artificial Intelligence Minister Evan Solomon is working on the government’s AI strategy, which will include support for Canada’s sovereign AI industry and child protection measures.
