Question Period Note: AI in the Public Service

How is the Government of Canada advancing its responsible AI adoption to improve how it serves Canadians?

• The Government is exploring areas where AI can improve service delivery and enhance the work of public servants.
• The AI Strategy for the Federal Public Service is guiding this work along with various policy tools that ensure the responsible use of AI.
• To provide transparency, we maintain a publicly available AI Register that lists where and how federal institutions are using or testing AI.
If pressed on the use of AI for ATIP tools:
• When it comes to the processing of ATIP requests, regardless of any technology used to support this work, all files are reviewed by ATIP professionals to ensure legal requirements for disclosure are being properly met.

The AI Strategy for the Federal Public Service, published in March 2025, is intended to accelerate the responsible adoption of AI across the Government of Canada to deliver better and more efficient services to Canadians, improve decision-making, advance research, and protect Canada and its interests.
TBS broadly engaged stakeholders, partners and the public to develop the Strategy, publishing a What We Heard Report in January 2025. Participants agreed that AI should be used to modernize and streamline service delivery.
The Strategy sets out key actions to advance responsible AI adoption within the Government of Canada. Some that contribute directly to public service modernization include:

• Establishing an AI centre of expertise to support teams developing and scaling AI tools
• Removing policy and legislative barriers and embedding a “think AI” approach within all planning and funding requests
• Providing all departments with common access to infrastructure, tools, and solutions
• Addressing obstacles to recruiting and retaining AI and data talent and upskilling current public servants at all levels to work with AI
• Outlining key activities for new and emerging digital roles and linking them to learning programs and paths to build future-ready digital professionals and teams

While AI use in the GC is not new, current and significant modernization projects are underway across the GC that leverage AI to drive better service, both internal and external to the public service. Examples include:

· The Benefits Delivery Modernization program (ESDC) is using AI to address:
o Workloads and backlogs for Employment Insurance claim recalculations
o Identifying potential cases of misuse or abuse of the Old Age Security program and fraud in temporary resident applications
· IRCC is using AI to:
o Complete administrative tasks for immigration applications, such as data matching and bulk email creation
o Streamline the processing of routine, low-risk cases for temporary work permits, passports, and temporary residency applications
· Launched in September 2025, GCTranslate is the first lighthouse project under the Strategy and stands out as a transformative solution for bilingual service delivery. Built on Canadian-trained AI models, GCTranslate is already improving translation speed and accuracy across departments, reducing turnaround times for official documents while maintaining linguistic quality

AI Use in Access to Information (ATI)
· TBS is committed to enhancing the ability of the public service to quickly and accurately respond to Access to Information and Privacy (ATIP) requests, while ensuring that the Access to Information Act (ATIA) legislative framework remains modern and consistent with jurisdictional approaches including meeting the expectations of Indigenous partners.
· Considering the ongoing workload pressure and volume of records being processed, there is an increased need to integrate automated solutions, including the use of AI, to support all aspects of request processing.
· To support ongoing and incremental improvement, TBS officials have been identifying innovative practices and tools, working with key partners to determine the scalability of solutions as well as engaging other departments as relevant to develop pilot projects that can be monitored and assessed. In some cases, TBS has engaged specific departments that are facing challenges (for example, National Defence and Library and Archives Canada).
· The pilot initiatives will serve to assess approaches that have proven to increase efficiency or have the potential to do so and facilitate a broader adoption by the ATIP community. TBS, as the policy center for ATI, is playing a coordination role in promoting the adoption of effective practices and providing additional support through guidance which simplifies the integration of such approaches.
· TBS has not adopted any AI tools or software to identify, suggest, or apply redactions in records processed under the ATIA.
· While other departments have explored AI-assisted redaction tools—such as National Defence, Public Services and Procurement Canada, and Transport Canada—TBS has not implemented such tools and AI tools are not used to make decisions or apply exemptions under the ATIA.

AI Project with the Canada School of Public Service (CSPS)
· A recent Hill Times article referenced several Government of Canada AI initiatives related to ATI. Key clarifications are:
o The project undertaken by TBS and CSPS did not involve automated record retrieval or de‑duplication, nor did it perform redactions. The initiative focused on encoding the ATIA into a machine‑readable format using Blawx, an open‑source tool developed by Lexpedite Technologies.
o The purpose of the project is to create a structured digital expression of the ATIA to support clearer and more consistent interpretation across institutions. The original cost of $90,000 was to cover CSPS resources; but these costs were ultimately absorbed by CSPS.
o The project has continued beyond 2024, with leadership now transitioned to the Canadian Digital Service (CDS) in partnership with TBS and Justice Canada.
o No additional funding has been provided to TBS, CSPS, or CDS.
· TBS continues to coordinate pilot initiatives while emphasizing compliance with the Directive on Automated Decision-Making and principles of transparency, fairness, and human oversight.

None