Canada Reviews AI Use in National Security Operations

The National Security and Intelligence Review Agency (NSIRA) of Canada is currently conducting a thorough examination of how artificial intelligence (AI) is utilized within national security operations. This review aims to assess the governance and application of AI technologies by security agencies, including their definitions, methodologies, and oversight mechanisms.

In a communication addressed to federal ministers and national security organizations, NSIRA chair Marie Deschamps emphasized that the findings from this study will provide valuable insights into the usage of AI tools. The goal is to identify any potential gaps or risks that may require further attention. Canadian agencies have employed AI for various tasks, from document translation to the detection of malware threats.

The review agency maintains a statutory right to access all relevant information held by the departments and agencies involved, including classified materials, with the exception of cabinet confidences. The letter, which has been made public on NSIRA’s website, indicates that data collection for the study may involve requests for documents, written explanations, interviews, and system access. Deschamps noted that the review may also include independent inspections of certain technical systems.

Key federal figures have received this correspondence, including Prime Minister Mark Carney, Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, Gary Anandasangaree, Minister of Public Safety, David McGuinty, Minister of Defence, Anita Anand, Minister of Foreign Affairs, and Mélanie Joly, Minister of Industry. Furthermore, it has been sent to leaders of major security agencies, such as the Canadian Security Intelligence Service (CSIS), the Royal Canadian Mounted Police (RCMP), and the Communications Security Establishment (CSE).

In response to inquiries about the review, the RCMP expressed its commitment to independent evaluations of national security activities. In a media statement, the RCMP highlighted the importance of establishing transparent external review processes to maintain public trust and confidence.

In 2024, a report from the National Security Transparency Advisory Group urged Canadian security agencies to provide detailed descriptions of their current and planned uses of AI systems. The advisory group predicted an increasing reliance on AI technologies for analyzing vast amounts of data, recognizing patterns, and interpreting trends. While CSIS and CSE acknowledged the need for transparency, they also noted limitations regarding public disclosures due to their security mandates.

The federal government has outlined principles for AI use, which include fostering transparency about AI applications and managing any associated risks to legal rights and democratic norms. Additionally, training for public officials involved in AI development is advocated to enhance their understanding of legal, ethical, and operational issues.

In its latest annual report, CSIS revealed that it is implementing AI pilot programs throughout the agency, adhering to the federal government’s guiding principles. Meanwhile, the RCMP’s official website outlines several critical factors in ensuring that AI is utilized legally and ethically. These factors include careful system design to prevent bias, respect for privacy in information analysis, and accountability measures to guarantee proper functioning.

The Communications Security Establishment has articulated its strategy for AI, committing to developing innovative capabilities that tackle significant challenges through responsible use of AI and machine learning. CSE’s strategy notes that, when deployed securely and effectively, these technologies could enhance data analysis speed and precision, thereby improving decision-making quality.

Caroline Xavier, chief of CSE, emphasized the organization’s commitment to a thoughtful and principled approach to AI adoption, ensuring accountability and responsibility remain central to its efforts. CSE acknowledges the inherent fallibility of AI technologies and is dedicated to rigorous testing and evaluation, keeping expert human oversight in the decision-making loop.

This ongoing review marks a pivotal moment for Canada’s approach to AI, underscoring the balance between leveraging advanced technologies and ensuring accountability within national security frameworks.