1. Six priorities for a Canadian and Québec strategy
Priority 1 — Make training available to workers already in employment
Problem: Most public discussion focuses on unemployed workers or people entering technology occupations. Yet the largest population affected by AI consists of people already employed in occupations being transformed.
The Job Market Monitor material offers several useful observations:
-
Most AI-exposed Canadian workers do not need specialized AI skills.
-
AI changes the competencies required in existing occupations.
-
Canadian employer-sponsored training is relatively limited and unevenly distributed.
-
Workers in smaller firms, lower-skilled occupations and less secure employment are particularly likely to have fewer training opportunities.
Job Market Monitor+1
Policy proposals:
-
Expand access to continuing education for employed workers.
-
Support training during paid working hours.
-
Provide financial incentives to employers who train existing employees.
-
Give priority to workers whose tasks are undergoing significant change, not just those facing imminent layoffs.
Priority 2 — Give employers and unions a central role in determining training needs
Problem: Employers understand the technologies they intend to introduce, while workers understand how their jobs are actually performed. Neither perspective alone is sufficient.
Policy proposals:
-
Establish joint employer–worker committees to assess AI-related changes in occupations.
-
Involve unions in identifying training priorities and selecting appropriate programs.
-
Conduct workplace assessments before implementing major technological changes.
-
Make training part of technological investment and organizational restructuring plans.
This is particularly important in Québec, where sectoral committees, collective bargaining and joint training initiatives can provide institutional mechanisms for anticipating skill requirements.
Priority 3 — Recognize transferable skills and competencies
Problem: AI changes the combination of skills required in a job. Traditional occupational classifications and formal credentials may not adequately reflect workers’ existing competencies.
Research on Canadian job postings highlights the importance of teamwork, communication, customer service and other social-emotional skills.
Policy proposals:
-
Develop competency-based assessment and recognition systems.
-
Recognize skills acquired informally through work experience.
-
Facilitate recognition of prior learning.
-
Develop short, modular training programs leading to recognized credentials.
-
Encourage training that combines technical, digital, interpersonal and problem-solving competencies.
Priority 4 — Make Employment Insurance a more effective instrument for training employed workers
This is arguably the most important institutional issue for Canadian labour-market policy.
Employment Insurance Part II already provides a framework for supporting skills development and employment-related training.
A stronger approach would recognize that technological transitions require training before workers lose their jobs.
Policy proposals:
-
Expand support for training that helps workers retain their existing employment as tasks evolve.
-
Support employer–union training partnerships.
-
Finance training for workers in occupations with high AI exposure, even when they remain employed.
-
Recognize formal, non-formal and workplace-based learning.
-
Measure success through employment retention, wage progression, improved competencies and productivity—not simply the number of people completing courses.
The policy distinction is important: training should help people keep an employment, not merely find another one after displacement.
Priority 5 — Ensure that productivity gains benefit workers as well as employers
Problem: AI can increase productivity, but the distribution of the resulting economic gains is not predetermined.
The Canadian research suggests substantial potential economic benefits from generative AI, but productivity improvements depend on actual adoption and organizational change.
Policy proposals:
-
Encourage employee participation in AI implementation.
-
Link public financial support for technological adoption to workforce development.
-
Monitor changes in wages, working conditions and job quality.
-
Promote opportunities for workers to move into higher-value tasks.
-
Encourage collective agreements to address the effects of AI on work organization.
Priority 6 — Develop better labour-market intelligence
Problem: Policymakers often have information about emerging technology occupations but much less information about how AI changes the tasks and skills required in ordinary occupations.
Policy proposals:
-
Track changes in occupational tasks, not just occupational employment.
-
Use online job postings to identify changing skill requirements.
-
Survey employers and workers about AI adoption and training needs.
-
Monitor training participation by age, education, occupation, industry and firm size.
-
Evaluate whether training actually improves employment outcomes.
This would also help distinguish genuine skill shortages from temporary employer preferences or exaggerated claims about AI-related skills.
2. Proposed implementation framework
The six priorities could be translated into a practical policy agenda:
| Area of intervention | Proposed instrument | Possible performance indicator |
|---|---|---|
| Training employed workers | Wage and training subsidies | Participation rate of employed workers |
| Employer–union cooperation | Joint training committees | Share of workplaces with joint plans |
| Transferable skills | Competency assessment and recognition | Skills formally recognized |
| Employment Insurance Part II | Expanded support for employed workers | Employment retention after training |
| Productivity sharing | Conditional innovation assistance | Productivity and wage growth |
| Labour-market intelligence | Occupational and task monitoring | Time required to identify emerging skill gaps |
3. Five particularly useful research contributions
These are the publications I would place at the centre of a policy paper.
1. AI and the changing demand for skills in Canada (2025)
Establishes why most AI-exposed workers need changing general competencies rather than specialized AI qualifications.
OECD · Canada · Skills demand
2. Green and AI Transitions – Training supply (2025)
Identifies the gap between the need for workforce training and the actual availability of appropriate courses.
OECD · Adult learning · Training supply
3. AI in the Workplace – What do workers and employers think? (2023)
Provides an important foundation for worker participation, workplace consultation and the quality of employment.
OECD · Workplace relations
4. Employer-Sponsored Training in Canada (2023)
Documents weaknesses in employer investment and unequal access to training.
Labour Market Information Council · Canada
5. AI and Skills – High-exposure occupations (2024)
Demonstrates that AI affects a much broader range of occupations than those requiring specialized AI expertise.
OECD · Occupational transformation
4. An important distinction for Employment Insurance policy
I would make a distinction between three types of intervention:
A. Preventive training
Workers remain employed but need to adapt to changing tasks and technologies.
B. Transitional training
Workers need to move to another occupation or employer because their existing employment is changing substantially.
C. Re-employment training
Workers have already lost their jobs and need help returning to employment.
My assessment is that Canadian employment policy should give considerably more attention to the first category.
Waiting for displacement before providing assistance is often economically inefficient and socially costly.
5. A possible central thesis for a policy paper
Artificial Intelligence and Workforce Development: From Job Displacement to Employment Retention
Artificial intelligence is transforming the nature of work, but its principal labour-market effect may not be the elimination of entire occupations. Rather, it is the reorganization of tasks, the changing composition of skills and the transformation of existing jobs.
This distinction has important implications for employment and training policy in Canada.
A policy approach centred primarily on displaced workers and the acquisition of specialized AI skills risks overlooking the much larger population of employed workers whose jobs are being transformed.
The priority should therefore be to strengthen the capacity of workers and employers to adapt to technological change while maintaining employment relationships.
This requires a more integrated workforce development strategy built around five principles:
- Prevention rather than remediation: provide training before technological change results in job displacement.
- Shared responsibility: give employers and workers, including their unions, a meaningful role in identifying training needs and determining priorities.
- Recognition of competencies: recognize transferable, informal and workplace-acquired skills alongside formal qualifications.
- Employment retention: make training accessible to employed workers through appropriate employment insurance and workforce development mechanisms.
- Shared productivity gains: ensure that technological innovation contributes not only to business competitiveness but also to improved employment quality, wages and working conditions.
The objective should not simply be to prepare workers for the jobs of tomorrow. It should be to enable workers to retain and improve their employment as the jobs of today evolve.
6. Suggested next research priority
Can Employment Insurance Part II be adapted to finance preventive training for employed workers whose jobs are being transformed by AI and other technological changes?
This would allow a focused examination of:
-
The interpretation of section 59 of the Employment Insurance Act, particularly training that helps participants keep an employment.
-
The respective roles of employers, unions and governments in determining training priorities.
-
The possibility of supporting formal and informal training.
-
The recognition of transferable competencies.
-
The use of sectoral committees and joint employer–union initiatives.
-
The measurement of employment retention and productivity outcomes.



Discussion
No comments yet.