‘We cannot wait to discover harms after the fact — strong oversight, proactive safeguards, and clear accountability are essential’
Canada's AI for All strategy has drawn praise from the business and research communities, but it is also generating pointed demands from human rights and labour voices who want the strategy's lofty equity commitments backed by enforceable accountability — particularly as the technology moves deeper into hiring, performance management, and workplace decision-making.
The Canadian Human Rights Commission issued a formal response to the strategy this week, with chief commissioner Charlotte-Anne Malischewski, accessibility commissioner Christopher Sutton, and pay equity commissioner Lori Straznicky issuing a joint statement that welcomed the document's ambitions while drawing a clear line on implementation.
"Artificial intelligence must advance, not undermine, the enjoyment of human rights in Canada," said Malischewski. "We cannot wait to discover harms after the fact — strong oversight, proactive safeguards, and clear accountability are essential. Where harm from AI systems does occur, people must have access to effective redress mechanisms."
The joint statement calls for human rights, accessibility, privacy, and equity considerations to be embedded throughout the entire lifecycle of AI systems — from procurement and design through to deployment and ongoing evaluation.
It specifically flags employment as a high-risk area requiring "strong oversight, clear regulatory alignment, proactive safeguards, and human rights impact assessment" before systems are deployed, not after.
Ottawa launched AI for All: Canada's new national artificial intelligence strategy last week, pledging more than $2.3 billion and targeting an additional $200 billion in economic growth to reshape how the country works and competes.
Workplace bias problem
The Commission's core concern is one HR professionals will recognize: AI systems trained on historical data can encode and amplify existing patterns of discrimination. The commissioners stated that "where existing systemic discrimination, bias, and exclusion are embedded in data and systems design, AI systems replicate inequalities and existing barriers and excludes diverse communities."
They called out Indigenous, Black, and racialized workers, people with disabilities, and women in all their diversity as groups most exposed to this risk.
"As we integrate AI across public and private sector systems and workplaces, we must ensure these technologies do not replicate or deepen existing systemic inequalities for women, in all of their diversity,” said Straznicky, calling for transparent design, inclusive data practices, and ongoing oversight as baseline requirements for AI systems that affect workers.
People with disabilities must have a meaningful role in shaping how AI tools are designed and governed, not simply benefit from them after the fact, said Sutton: "Accessibility must be built in by design, not added as an afterthought.”
The Commission also raised the issue of redress — a gap that is often missing from employer AI governance frameworks. The statement emphasizes that where AI causes harm, accessible and effective complaint and correction mechanisms must already be in place.
The federal government also committed to fast-tracking the immigration of skilled artificial-intelligence workers, and to retaining them once they arrive, under the new national AI strategy.
Labour's warning: No rules, no timelines
Canada's largest union of public sector workers went further, calling the strategy fundamentally incomplete.
The National Union of Public and General Employees (NUPGE) said the strategy "fails to address many of the serious problems associated with AI," and warned that without action, AI will be used in ways that harm Canadians while the only winners are private equity firms "looking to make a quick buck and unconcerned about the long-term, well-being of workers, the economy and the environment."
"The strong legislation and regulation needed to protect people and the environment from the potential abuses of AI are missing from the federal government's strategy. Without effective and enforceable rules and public control over key infrastructure, the AI strategy is like playing short-handed in hockey,” said NUPGE president Bert Blundon.
Two specific gaps drew NUPGE's attention. First, the strategy mentions plans for legislation in several areas, but there are no timelines for when legislation will be introduced and no details of what will be in the legislation. Second: there is nothing about strengthening labour legislation in the AI strategy, and no proposal to strengthen employment standards legislation to protect people from employers using AI to spy on them or for other violations of their rights.
The union also raised a concern with practical relevance for any employer handling sensitive data: the US CLOUD Act allows the US government to force companies with a US presence, even if those companies are Canadian-owned, to produce data stored on servers outside the United States — meaning a Canadian company with US operations could be forced to provide data it stores in Canada to American authorities.
Workforce alliances for different sectors
The strategy itself does include a significant pro-worker architecture that employers will be asked to participate in directly. The document commits to launching sector-specific Workforce Alliances across six priority industries, designed to bring together employers, governments, unions, post-secondary institutions, and Indigenous partners to map skills gaps, coordinate training pipelines, and align investment with labour market transformation.
The strategy frames it as a mechanism for employers to help shape how AI-driven disruption is managed — with the explicit goal of making adoption "pro-worker, practical, and productivity-enhancing." Employers in healthcare, energy, transportation, agriculture, and manufacturing are the most likely first movers, given the strategy's five designated priority sectors.
The strategy also commits to targeted programs for "female-dominated sectors most exposed to early disruption," delivered through the existing Women's Program.
