Part 4 of 8: AI for All Pillar 2 Review: Empowering Canadians—The Human Core of the Federal AI Strategy

Nikki Matarazzo, AI Evangelist

July 20th, 2026

A national technology strategy is only as robust as the people tasked with using it. In its second major pillar, Empowering Canadians, the AI for All strategy directly tackles the most volatile and human element of the artificial intelligence boom: our workforce and our communities. Anchored by the interconnected priorities of literacy, opportunity, and participation, this section of the strategy aims to transition the public from passive consumers to active shapers of an AI-driven economy. It is a noble and necessary vision. A tool is only as useful as the person using it. However, as the strategy moves from abstract ideals to multi-million-dollar implementation, a closer look reveals where the foundation of the pillar is strong, and where there are cracks in the road to empowerment.

The strategy gets several critical things right, starting with its pragmatic approach to public infrastructure. By launching the National AI Literacy Initiative through public libraries and community hubs, the government avoids treating AI as an elite technical skill confined to urban tech corridors or university labs. Instead, it leverages existing, trusted public institutions to reach rural, remote, and northern regions. This grassroots approach actively works to prevent a geographic digital divide from hardening before it even starts, ensuring that baseline technical literacy is universally accessible.

Furthermore, the strategy avoids the common trap of treating the labor market as a monolith by identifying and safeguarding specific disruption zones. Acknowledging that female-dominated sectors face distinct early disruption shocks, the strategy integrates AI upskilling into the existing Women’s Program while concurrently addressing the rise of AI-fueled online abuse. Similarly, the focus on the blue-collar transition—demonstrated by funding initiatives like Amii’s AI Pathways program for energy workers and modernizing the Job Bank with a fifty-million-dollar AI-powered upgrade—shows a sophisticated understanding that mid-career, industrial, and trade workers require immediate, targeted intervention rather than generic advice.

Perhaps the most unique strength of this pillar is its treatment of cultural identity and sovereignty as technical requirements rather than afterthoughts. The inclusion of dedicated funding for Indigenous-led AI initiatives in partnership with organizations like Mila, alongside the establishment of the fifty-million-dollar Creative Technology Program, represents a major policy win. Because large language models and creative tools default to the data they are trained on, they are overwhelmingly biased toward American and Eurocentric perspectives. By actively funding local language models, land management systems, and cultural preservation tools, Canada is ensuring that its diverse multicultural identity and official languages are built into the very architecture of the tools we use.

Despite these commendable ambitions, there are a few omissions and logistical assumptions that threaten to stall the strategy's real-world impact. The first major blind spot is the platform sovereignty paradox introduced by the promise of providing all post-secondary students with access to trusted AI agents. The strategy fails to clarify who is building, hosting, and securing these systems. Currently, the raw infrastructure and compute power required to run highly capable AI agents at scale are dominated by a handful of American big-tech conglomerates. If Canada builds its foundational student literacy on top of proprietary, foreign-hosted platforms, it creates a systemic dependency that compromises data sovereignty and exposes student data to foreign corporate terms. The questions we need to ask are: which agents are students going to be using? What is the cultural identify of said agents? How are teachers, administration and students addressing agent bias?

Pillar two also suffers from a significant scale-versus-depth dilemma within its educational targets. The framework outlines an ambitious goal to reach one million post-secondary students while training only three thousand educators with AI learning kits. This ratio represents an extreme bottleneck that undercuts the efficacy of the initiative. Furthermore, relying on classroom learning kits is a surface-level patch for a deeply systemic educational issue. Without deep, structural integration into provincial teacher-college curriculums and sustained professional development, these literacy initiatives risk becoming superficial, one-off modules that fail to build true critical thinking or technical capability.

There is also an economic void regarding enterprise adoption. While the government projects the creation of over two hundred and fifty thousand new AI-relevant jobs by 2031 and is heavily funding student placements, it overlooks the fact that jobs are created by businesses, not training programs. Historically, Canadian small-and-medium enterprises have been notoriously slow to adopt digital technologies and innovate at scale. By focusing almost exclusively on supplying AI-ready workers without offering aggressive, direct economic incentives for domestic companies to restructure their workflows and hire them, we run a massive risk. Canada may end up training a brilliant generation of AI-literate workers who are ultimately forced to take their talents to foreign employers because local businesses simply lag too far behind to utilize them. Think about it. How many Canadian small and medium size businesses have a conversational chat agent on their web pages today?

To bridge these gaps, Canadian businesses and community leaders cannot afford to wait around for federal funding to trickle down; they need to take the reins on adoption. Organizations should proactively establish internal "AI sandboxes" and joint worker-management committees to experiment safely with tools, ensuring that employees are co-designing how AI optimizes their day-to-day tasks rather than having it forced upon them. By partnering directly with local polytechnics and colleges for custom, sector-specific micro-credentials, businesses can bypass bureaucratic bottlenecks and build the exact skills their specific regional workforce needs. This grassroots demand forces enterprise adoption from the bottom up, creating immediate, localized roles for the very workers the government is trying to train.

On an individual level, Canadians need to shift from being passive users of foreign tools to active creators and critical evaluators within their own circles. We can plug the literacy gaps by organizing informal peer-learning circles in our workplaces, schools, and local community groups to share open-source alternatives and interrogate the biases embedded in commercial AI outputs. Educators and professionals can take ownership of their data sovereignty by championing local, open-source models over proprietary corporate software, ensuring that our collective data stays within our control. By demanding transparency from employers and local institutions regarding how our data is handled, everyday citizens can actively construct the human-centric guardrails required.

Ultimately, the Empowering Canadians pillar successfully re-centers the national AI conversation around human dignity, labor rights, and cultural inclusion. It correctly asserts that a society cannot thrive if its citizens feel alienated by progress. We do need, however, a stronger focus on the infrastructure side of sovereign compute and enterprise adoption. If Canada wants to truly empower its citizens, we must bridge the gap between teaching people how to use AI and building the domestic economic engine that allows them to do so.

Until next time …