Madam Chair, I'm talking about AI, so this is on topic.
Here's the second recommendation:
The Federal Government should build a public AI stack which includes cloud infrastructure, data management and AI models and applications. This public physical and digital infrastructure would promote accountability and transparency, advance the public interest and develop Canada's digital sovereignty.
The other thing, which is in the third recommendation, is this:
The Federal Government should operate under an accountability framework that centers consultations with and input from unions, academic researchers, civil society, among others, and ensure all decisions are based on unbiased research and data mined from sectoral tables and CUPE's proposed Innovation Hub and Canadian Observatory on AI.
There's a fourth recommendation:
The Federal Government should invest in foundational digital literacy education for Canadians, in training and re-training public sector workers who are or will interact with AI and in social safety net supports such as Employment Insurance for workers who will be displaced because of AI.
Among the additional questions that CUPE would have liked the government to incorporate into its strategy is one about research needed to understand the impact of AI on work:
In Canada, we do not yet have a complete picture of AI's impact on working conditions, job loss, and skills and training needs of workers. We do know that AI can affect different job classifications in disparate and distinct ways – even when they are in the same workplace. CUPE's own research and consultations with members have found that AI's disparate impacts are best captured through a combination of qualitative and quantitative methods. Internationally, we have varying assessments of how AI will affect jobs. Most economists say the impact is hard to predict because we don't know exactly how AI will develop and how quickly. Statistics Canada has found that 31% of workers are in jobs likely to dramatically change or disappear because of AI and 29% of workers are in jobs where they will likely do their jobs alongside AI systems. Some research has indicated that the impact of AI implementation will be gendered with twice as many women-dominated jobs at risk compared to men-dominated jobs. Research is also showing how increasing AI-amplified monitoring, surveillance and algorithmic management in the workforce is affecting the well-being of workers, the human rights of workers and power dynamics between workers and employers.
The brief includes a footnote here. It's important to mention the research consulted. This information is from a scientific article by Antonio Aloisi and Valerio De Stefano, accessed in 2025, entitled “AI at Work, Algorithmic Bosses, and the Ambivalence of Automation”, which was published in Artificial Intelligence and Labour Law. It's also from an article by Armin Granulo et al. entitled “The Social Cost of Algorithmic Management”, which was published in the illustrious Harvard Business Review. The brief also cites an article by Benta Mbare et al. entitled “Algorithmic Management, Wellbeing and Platform Work: Understanding the Psychosocial Risks and Experiences of Food Couriers in Finland,” as well as an article entitled “Algorithmic Discrimination: Examining Its Types and Regulatory Measures with Emphasis on US Legal Practices”, both published in Frontiers in Artificial Intelligence.
As you can see, I'm practising my English. Some Quebec Liberal MPs are in the habit of speaking English often, and although they really inspire me, I am acutely aware that Quebec's only official language is French. Nonetheless, that doesn't stop us from knowing how to speak another language. Indeed, we know that people in Quebec are the most bilingual even though Quebec has only one official language: French.
I would like to continue my presentation about CUPE's submission. I was talking about the tools that, for example—