The government's current patchwork approach to AI in the federal public sector is inadequate to keep pace with rapid technological changes. PIPSC remains disappointed by the TBS strategy unveiled in March 2025.
We’ve long advocated for a federal strategy that guarantees transparency, including a registry on all AI use – a commitment previously made by Chief Data Officer of Canada, Steven Burt—but this was not delivered. There were also no assurances against job cuts due to AI or against its misuse from monitoring public servants, which we know is already happening.
Most critically, there was no mention of an independent AI regulator. AI, when used responsibly, can empower workers, enhance jobs, and drive efficiencies. However, its unregulated use invites abuse: replacing workers, threatening privacy and undermining public sector standards.
PIPSC wants to see more substantive engagement and ongoing consultation on AI, and we continue to call for the creation of joint federal union and government committees. Public servants deserve to understand how AI will reshape their jobs and workplace, and our members deserve a stronger voice in this transformation. It's too crucial for the future of the public service for workers to be sidelined or their critical feedback ignored.
Canada’s federal public sector needs a comprehensive and transparent approach to AI —supported by regulation and independent oversight, ongoing consultation, and collective bargaining— that addresses privacy risks, ensures employee rights, and safeguards good public sector jobs.
Specifically, PIPSC is advocating for a federal strategy on AI that includes the following protections, safeguards and principles:
- Consultation: Whether it’s introducing new AI into workflows or developing a federal strategy, government must consult unions and affected parties at all times. Establish regular consultation mechanisms across sectors, including with the Treasury Board, meeting at least four times annually. Strengthen the AI Advisory Council by incorporating union representation as we are directly affected by AI use.
- Regulation: Establish federal oversight bodies to regulate AI across all levels of government, including national security institutions, ensuring privacy and human rights protections. Ideally, Canada establishes a single regulatory framework that governs AI in both private and public sectors.
- Transparency: Treasury Board, departments, and agencies must publicly disclose all AI uses, particularly regarding recruitment, staffing, performance management, security screning, and visitor access.
- Equity, Diversity, and Inclusion (EDI): AI policies must be developed with an EDI lens, ensuring no discriminatory practices, like Clearview photo recognition, are used. All AI must be inclusive from the outset.
- Bargaining AI: Mandate the inclusion of AI use in the collective bargaining process so that employers and unions can bargain on how AI affects an employee’s work life.
- Monitoring: End all covert AI employee monitoring, including surveillance tools and robots, as highlighted by recent media reports.
- Review: New AI systems must be regularly reviewed after implementation to assess their effectiveness and make necessary adjustments.
- Whistleblower Protection: Protect employees who report unethical AI use, following successful models like those in Australia.
- Research: Federal departments should lead AI research to enhance existing programs or develop new ones, ensuring that research is conducted by public servants unless exceptions are justified.
- Training and Retraining: Provide mandatory, fully-funded AI training or retraining for employees, starting immediately when needed.
- Advance Warning: Extend the notice period for AI-related job changes or layoffs to at least one year, allowing sufficient time for retraining.
- Job Guarantee: If retraining is impractical, guarantee new job opportunities for employees impacted by AI-driven layoffs.