Source: CTV News

CTV News reports that Canada’s National Security and Intelligence Review Agency (NSIRA) has initiated a comprehensive study into how national security organizations utilize and oversee artificial intelligence. The review focuses on agencies such as the Canadian Security Intelligence Service (CSIS) and the Royal Canadian Mounted Police (RCMP), examining their pilot projects for transcribing audio, translating materials, and summarizing complex documents. The watchdog aims to identify potential gaps in governance and assess risks to legal rights and democratic norms posed by the fallibility of these emerging technologies.

Federal principles for AI use in Canada emphasize openness, transparency, and the maintenance of "human-in-the-loop" protocols to ensure accuracy. CSIS has noted that all AI outputs undergo rigorous human review and algorithmic impact assessments to manage negative effects. The NSIRA study comes as national security agencies increasingly rely on AI to analyze vast volumes of text and recognize behavioral patterns. By evaluating current practices, the watchdog intends to provide insights that will guide future policy and ensure that the adoption of AI in intelligence gathering remains lawful and ethically responsible.