NorTech Innovations & Solutions

Canada’s 2025 Budget, tabled on November 4, marks an ambitious shift in public sector operations, positioning Artificial Intelligence (AI) as a central driver for productivity. Backed by a $925.6 million investment in sovereign AI infrastructure and the establishment of the TechStat program to measure its national impact, the message is clear: Canada is committing to the AI race.

Amidst a significant push to reduce the federal public service by approximately 40,000 positions by 2028-29, the budget explicitly links technological integration—specifically AI—with the creation of a “tech-enabled public service of the future.”

The government’s plan, detailed in sections like “Optimizing Productivity in Government” and “Adopting AI to Enhance Productivity and Improve Services,” is a clear move to modernize and achieve an audacious savings plan. However, this transformative vision, while promising increased efficiency, inevitably raises complex questions that demand a deeper look—namely around information security, ethical governance, and the future of the Canadian workforce.

The Tech-Enabled Public Service: A Double-Edged Sword

The budget’s commitment to adopting AI is not abstract; it points to concrete applications across key departments:

• Shared Services Canada (SSC): Will apply AI and automation to streamline internal operations and automate common IT support requests, aiming to reduce call volumes and costs.
• Department of Justice: Will integrate AI and advanced analytics to streamline routine tasks and enhance decision-making, ostensibly freeing employees for “higher-value strategic work.”
• Transport Canada: Will use AI and automation to optimize back-office activities and realign service delivery models with a focus on self-service solutions.

This push, overseen by a new Office of Digital Transformation, is designed to “proactively identify, implement, and scale technology solutions” and to “leverage expertise from internal sources and the private sector to hasten AI adoption.”

The Outsourcing and Security Tightrope

The budget’s mention of leveraging private sector expertise immediately brings information security concerns into sharp focus. Major AI projects often involve significant outsourcing, meaning sensitive government data, including Canadians’ personal information, will be processed or stored in non-Government of Canada (GC) systems.

• The Data Custody Challenge: While the Canadian Centre for Cyber Security provides guidance for protecting specified information in non-GC systems (ITSP.10.171), the sheer volume and sensitivity of data processed by new AI tools dramatically raise the stakes. Outsourcing creates a distributed security perimeter, requiring stringent contractual obligations and continuous auditing of private partners to ensure compliance with the Treasury Board’s policies on privacy and security. As the perimeter expands, so too does the complexity of securing the data and ensuring privacy compliance across multiple external partners, potentially leading to a more fragmented and harder-to-monitor security landscape.
• Digital Sovereignty: In a critical move, the budget announces a made-in-Canada AI tool to be developed by Shared Services Canada (SSC) in partnership with the Department of National Defence (DND) and the Communications Security Establishment (CSE). This explicit focus on protecting our digital sovereignty and keeping government data safe in Canada is a direct attempt to mitigate the risks of using foreign-based, proprietary AI models, offering a potential security advantage and opportunity for domestic tech companies.

While AI adoption has the potential to streamline government services, it also raises critical concerns related to outsourcing, security, and ethics—issues that are closely interlinked.

Relying on private-sector AI solutions means outsourcing critical government functions and storing sensitive data outside of government control, heightening the risk of breaches, data misuse, and loss of accountability.

AI’s Ethical Minefield: Bias, Justice, and Accountability

Perhaps the most significant long-term challenge lies in the ethical implementation of AI, particularly in sensitive areas like the Department of Justice. The inherent issue of algorithmic bias is not a hypothetical risk—it’s a well-documented phenomenon where AI systems, trained on historical data, can perpetuate and amplify existing societal biases, especially against marginalized communities.

The Justice System Conundrum

When AI tools are used to “enhance decision-making” in the Justice Department, even if only to streamline routine tasks, the integrity of the outputs is paramount.

• Bias and Fairness: If an AI system, for example, is trained on historical legal data that reflects systemic biases in arrests, sentencing, or case prioritization, its recommendations will likely reproduce those unfair outcomes. The lack of detail in the budget report regarding the ethical guardrails is a glaring vacuum that must be filled.
• Transparency and Accountability: In a justice or governmental context, citizens must have the right to know when an automated system has influenced a decision that affects their lives. The principles of transparency and explainability—understanding why the AI reached a particular conclusion—are non-negotiable for upholding procedural fairness and the rule of law. The federal government’s existing Directive on Automated Decision-Making is a starting point, but the rapid evolution of AI demands a much more robust and legally enforceable ethical framework.

The introduction of AI in the Justice Department calls for not just retrospective checks but continuous, preemptive audits to ensure these algorithms remain fair and free from bias throughout their deployment.

AI and the Canadian Workforce: The Attrition Question

The Budget 2025 explicitly projects a reduction in the public service population by approximately 40,000 positions (a 10% decline) by 2028-29, driven by attrition and the Comprehensive Expenditure Review. Simultaneously, it announces the introduction of “new and better tools such as AI” as the mechanism to ensure the public service “will still deliver the services that Canadians depend on.”

The juxtaposition is impossible to ignore: AI is being introduced as an efficiency and cost-saving tool at the same time a significant reduction in the workforce is being planned.

• Productivity vs. Displacement: While the government frames this as AI tools helping public servants focus on “higher-value strategic work,” public sector unions and labour advocates are concerned that AI will function primarily as a job replacement tool, weakening the social safety net and public service capacity.
• The Skills Gap: This transition requires a massive investment in upskilling and reskilling the remaining public servants. The budget mentions training to equip public servants, but the success of the “tech-enabled public service” hinges on whether the government can truly retrain its workforce to manage, audit, and work alongside AI systems, shifting from manual processing to sophisticated data governance and ethical oversight. The risk is that the skills gap widens, leading to even more reliance on expensive private sector consultants. If not properly addressed, the skills gap may lead to even greater reliance on expensive private sector consultants, which could undermine the government’s goal of creating a more efficient public service.

While the promise of AI is efficiency and innovation, it is essential to consider the human cost. The 40,000 job cuts may disproportionately affect lower-income public service workers, potentially deepening social inequality.

What happens when AI becomes more than just a tool, but a driving force behind the future of government work? Will those displaced workers find a place in the new digital world?

A Defining Moment for Canadian Tech Policy

Canada’s Budget 2025 presents a generational opportunity to harness AI for a more efficient and productive government. The commitment to a $1 trillion-plus investment target with a focus on AI, infrastructure, and R&D is a decisive economic strategy.

However, the implementation of AI across federal departments is where the rubber meets the road. The government must move past vague intentions and provide a detailed roadmap that explicitly addresses the following:

1. Strict Security Frameworks: Concrete, enforceable security and data sovereignty requirements for all private-sector AI partners, particularly those handling personal information. The more data is outsourced to private contractors, the harder it becomes to enforce Canada’s digital sovereignty, putting both security and privacy at risk.
2. Robust Ethical Governance: Specific, pre-emptive policies to audit and mitigate algorithmic bias, ensuring fairness and transparency, especially in the Justice Department.
3. Workforce Transition: A funded and comprehensive plan to retrain and upskill public servants, ensuring AI enhances human work rather than simply replacing it, thereby protecting essential services.

AI is poised to fundamentally reshape the Canadian public service. The challenge now is to ensure this transformation is not just fast and cheap, but secure, equitable, and accountable to all Canadians.

Food for Thought: What Does a Sovereign AI Future Mean for You?

The Budget 2025 asks Canadians to trust in a new, tech-driven era of government. This trust is earned not just through efficiency, but through responsible implementation. As these plans move from paper to practice, here are the key questions we should all be considering:

• Data Control: If the government moves services to a “made-in-Canada” Sovereign Cloud, does this new layer of data control make you feel more or less secure than relying on established global infrastructure providers?
• The Bias Check: Given the Department of Justice’s plan to use AI for “decision-making,” who should be tasked with regularly auditing these algorithms for bias, and should external, independent bodies be given the authority to oversee and intervene in AI decision-making processes?
• Your Data Trail: As the TechStat program tracks Canadians’ engagement with AI services, how much personal data should be collected to improve these services without infringing on privacy?

The Budget 2025 has laid the foundation. Now, it is up to Canadians and tech leaders to hold the government accountable. As AI fundamentally reshapes the public service, our shared challenge is to ensure this transformation is not just fast and cheap, but secure, equitable, and accountable to all Canadians. Will we be vigilant stewards of this digital future?