After an extensive public consultation process, the federal government released its new AI strategy for the public service last week. Here’s a summary of the main features.
The rationale for utilizing AI more within the public service is extensive and obvious. The strategy states:
“AI can streamline or automate routine tasks for public servants, freeing them to focus on more complex and critical work. It can increase the public service’s efficiency, effectiveness and productivity, maximizing its value to Canadians. It can improve the speed and scale of data and information analysis far beyond what was once possible, leading to faster, more informed decision making and scientific discovery. It can also create new avenues for public engagement…”
Vision and Principles
This is the vision statement for the AI strategy:
“By responsibly adopting AI, the Government of Canada can deliver world-class services to its clients, protect our people and interests, achieve a more innovative and efficient workplace, and accelerate scientific discovery for the benefit of all.”
There are four key principles underneath this:
- Human-centred
- Collaborative
- Ready
- Responsible
The scope of what is considered AI was also stipulated. It encompasses knowledge-based systems, machine-learning systems, and key application areas such as computer vision, natural language processing, speech recognition, intelligent decision support systems, and intelligent robotic systems.
The number of AI priority areas have been narrowed to four in the new strategy from the consultation document. They now include:
Priority 1: Central AI capacity
Priority 2: Policy, legislation and governance
Priority 3: Talent and training
Priority 4: Engagement, transparency, and value to Canadians
Each priority area is accompanied by three key actions. Let’s dig into each a bit.
Central AI Capacity
The strategy identifies the need for “a central hub to support project implementation and share knowledge” across the public service. To accomplish this, three key actions are listed:
- Establish an AI Centre of Expertise for the Government of Canada – to Identify high-value use cases for AI integration; support data readiness; support with procurement, governance, assessments, review processes, and other requirements. The AI Centre is also meant to be a convenor for government-wide knowledge sharing and act as an accelerator for AI solutions.
- Enable common infrastructure – providing for high-performance computing and cloud infrastructure that is available, secure, and scalable to meet the demands of AI projects; common data and information management systems and practices; to creating a standard GoC platform service for vendors.
- Identify and develop a lighthouse project – A test case to identify barriers and project teams’ needs for support during development and obstacles to subsequent customization and scaling; act as a pilot for the development of government-wide governance processes.
Policy and Governance
The strategy baldly states there is “no common model for AI governance yet exists within the Government of Canada.” Rather, a patchwork of rules exist instead. There is a need to create an up-to-date legislative and policy framework to govern the adoption and use of AI within the public service. These three key actions are listed:
- Establish common AI governance and risk management frameworks – to provide clear guidance to AI project teams; address potential risks associated with AI use, such as data privacy and security, bias detection and mitigation, model interpretability and explainability, environmental impact, and human involvement; lay out the security, trust, and reliability controls necessary for system resiliency; link with AI Safety Institute on ethics and governance.
- Address policy and legislative alignment, gaps, and barriers – clear away obstacles to AI adoption; clarify the responsibilities of chief information, data and privacy officers for AI adoption, and the acceptable use of AI by outside organizations in their interactions with the GoC; address legal and policy ambiguities related to privacy and training data and the application of the national security exemption; update procurement policies to be more responsive to AI requirements.
- Adopt a “think AI” approach – require departments to (a) identify three areas, programs or services with business problems that have a high potential to be solved using AI; (b) consider solutions and resourcing requirements for AI and its enablers at the outset of initiatives; (c) prioritize AI infrastructure and secure adoption in departmental integrated IT planning processes; (d) develop their own AI strategies to ensure alignment and effective use of resources.
Talent and Training
The strategy recognizes the current lack of qualified AI practitioners within the federal public service, citing “a 30% vacancy rate for digital roles”. Demand for AI expertise is exploding, and Canada has to actively recruit and retain qualified people if it wants to do AI well. These three key actions are listed:
- Develop a training plan – this will include both general and tailored training. General training will be directed at increasing understanding of and confidence in using AI, including embedded AI capabilities; developing skills associated with responsible, secure, and effective use, including effective prompt engineering; and establishing leadership programs to achieve a culture that promotes AI adoption. Tailored training will address both more specific and advanced technical skills, and the behavioural skills needed for successful adoption, such risk identification and management, and effective leadership of AI projects and teams.
- Benchmark talent needs – benchmark talent requirements for AI and its employees’ existing AI knowledge and skills across the enterprise and use these benchmarks to develop learner personas with accompanying training plans and identify employees who could be offered further training.
- Develop a talent plan – to recruit and retain talent. This plan will explore obstacles to recruitment and retention and ways to establish flexible data science career pathways for AI practitioners.
Engagement, accountability, and transparency, and value to Canadians
Consultations identified the need for actively and systematically engaging on AI, noting “levels of mistrust in AI and its use are high.” In particular, there is a need to be conscious of AI “on those more greatly impacted by algorithmic bias or barriers to access”. These three key actions are listed:
- Strengthen accountability and transparency on AI use – establish a public register of AI systems within government. The register will include information about what and how data is being used, how it was trained, and what quality assurance and privacy and security measures are in place.
- Demonstrate impact and value to Canadians – develop metrics and performance indicators to demonstrate the impact and value of AI initiatives to those we serve; track AI adoption across government.
- Commit to engagement on AI – conduct early and meaningful public and stakeholder engagement on AI initiatives of significant public interest or concern. This will include targeted engagement of communities that face greater impacts, risks or barriers from AI systems and union and employee engagement on workforce impacts.