Initiative overview
The new Air Travel Complaints regime, launched in 2023, was designed to modernise Canada’s air travel complaint process. In FY 2024‑25, the Canadian Transportation Agency (CTA) closed 33,000 complaints, surpassing its target, but annual intake continues to exceed 40,000, leaving a backlog of roughly 90,000 complaints.
To address this challenge, the CTA is rolling out AI‑enabled tools that streamline key workflow steps while preserving decision quality and consistency. The first tool supports quality review of drafted decisions, and the second introduces automated document summarisation along with integrated case‑file summaries. Future phases will explore AI‑assisted decision drafting, enhanced triage models, and potential external tools to help passengers' self‑screen their complaints.
To support the safe and responsible adoption of AI, the CTA has implemented an AI Governance Framework that provides structured oversight for all AI initiatives. The framework establishes clear expectations for risk management, transparency, and accountability, and incorporates mandatory requirements such as the Algorithmic Impact Assessment. This approach ensures alignment with federal direction, including the Directive on Automated Decision‑Making, and enables secure, privacy‑protective, and ethical deployment of AI across the organisation.
Results, outcomes and impact
CTA expects reduced time on task, more consistent decisions, and faster resolution of the 90,000‑complaint backlog.
Challenges and lessons learned
Lessons learned highlight that introducing AI into a high‑volume programme requires both substantial operational change and new organisational capacities. Teams must develop new skills such as prompt engineering while adapting workflows and ensuring that privacy, ethics, and transparency requirements remain at the forefront. Effective change‑management practices, including training, user‑support mechanisms, and clear communication of the costs and benefits of adopting the tool versus maintaining the status quo, are critical for adoption and for reducing resistance
Equally important was recognising that AI governance itself represents a major organisational change. Establishing a formal governance structure early clarified expectations for documentation, risk management, and completion of Algorithmic Impact Assessments. This foundation enabled responsible experimentation while ensuring alignment with federal requirements.
