Immigration lawyers say automation is partly driving a massive Federal Court backlog

Canada’s Federal Court immigration caseload has exploded, rising from 6,400 cases in 2020 to over 28,000 in 2025, with more than 6,600 filed in the first quarter of 2026 alone. Most cases involve non-refugee visa applications, and lawyers attribute the surge partly to Immigration, Refugees and Citizenship Canada’s (IRCC) increased reliance on artificial intelligence and automation for processing decisions. Immigration lawyer Jacqueline Bonisteel argues that AI-driven tools reduce human oversight, leading to generic refusal decisions lacking engagement with submitted evidence. She notes that five years ago, officers provided detailed explanations for rejections, whereas current automated systems often produce standardized, impersonal responses. IRCC’s press secretary, Taous Ait, countered that AI tools—used for triaging, summarizing, and chatbot responses—operate under human supervision and do not influence final decisions. The department’s AI strategy, released earlier this year, emphasizes using technology to expedite low-risk cases while ensuring all refusals are reviewed by trained officers. Despite IRCC’s assurances, lawyers report persistent errors, such as a visitor visa denial based on incorrect assumptions about family ties. Nalini Reddy, a Winnipeg-based lawyer, cited a case where a client’s application was rejected despite clear evidence her children and extended family resided in the Philippines, not Canada. The backlog and quality concerns highlight tensions between efficiency and accuracy in immigration processing. While IRCC insists on human oversight, critics argue that automation risks undermining fair decision-making in a system already strained by record caseloads.