Madam Chair, thank you for being so considerate of committee members and ensuring that they're in attendance when they ask to speak. I have noticed that that isn't the case for all members of this committee.
The brief I'm sharing goes on to say:
For example, AI hiring tools have been found to discriminate against women, Black and racialized people, older people and people with disabilities while AI surveillance and monitoring and automated decision-making tools in the workplace can intensify work, harm the health and safety of workers and erode the privacy rights of workers. These tools can discourage and penalize workers from taking breaks to which they are entitled, increase the speed of work and result in psychological harm and mental health problems for workers in cases of high demands and little control over the work process. Employment law used to […] for a reason. The balance of power shifts even further toward control and coercion when employers implement invasive AI algorithmic management and surveillance technology.
CUPE's brief also states:
We need to develop the data and research—both quantitative and qualitative—to understand these shifts in individual workplaces, sectors and the Canadian economy as a whole in order to understand the impact of AI on work. This data would support active government planning to address and mitigate negative impacts of AI on workers. Statistics Canada should collect data about the impact of AI on work and workers. This data should be supported by ongoing qualitative research funded by the federal government assessing the privacy, human rights, skills and training and physical and psychological health and safety impacts of AI on workers.
CUPE addresses another question: “How can Canada support workers affected by AI?”. Here's its answer:
A key question that workers will be looking for in the AI Strategy is whether their jobs are at risk of being eliminated and how they will be supported to learn new skills. They will be looking for the federal government to create and deepen social policy that ensures pathways and ladders to train and re-train in new jobs and to provide income support during transition periods. Workers should not pay the costs of technological changes. Governments must ensure our social safety net is strengthened to support this transformation through improved employment insurance benefits, targeted training and re-training programs and active labour market strategies. Workers also need governments to put in place laws and regulations to ensure AI does not result in the loss or deterioration of good jobs. The federal government should work with provincial governments to mandate transparency when AI systems are introduced in the workplace.
Transparency isn't something this government does.
Right now, workers are often not being informed about data collection or AI implementation in their workplace. Employers should be required to provide information to workers, unions and the public about the purpose of the system or data collection and the parameters, rules and instructions underlying the system. Federal and provincial laws should put in place restrictions and safeguards when algorithmic management systems are used, as these systems can lack transparency, explainability and fairness, be biased or discriminatory, and put the health and safety of workers at risk. Algorithmic management systems process employee data through opaque algorithms to hire, assign tasks, set schedules, assess performance, and recommend discipline or dismissal. Workers also need strong data protection and privacy laws that protect their dignity and autonomy in the workplace in an era of invasive bossware technology that can track their movement, emotions, tone of voice and output.
I will continue by citing another question raised in the same brief: “Who needs to be consulted in the AI Strategy who isn't already?” We cannot even know this, as members of Parliament. The government won't tell us who participated in the consultation or what was said. So we're left to rely on assumptions.
I'll read you the union's response:
The best way to build trust in AI is to ensure the breadth of voices that are affected by AI are involved in consultations on how Canada addresses AI. Though CUPE is pleased to have a staff member involved in the federal AI strategy task force, the overall representation skews toward industry and pro-industry researchers.
There is already a problem: a lack of representation within a group that the government itself created.
The union stated the following:
The federal government should ensure an array of unions are involved in consultations as workers in different sectors will be affected in disparate ways. For example, workers in the creative sectors are already being affected in distinct ways and require targeted measures to protect their rights and their jobs. A much wider array of civil society voices needs to be present, bringing unique knowledge on privacy, data rights, civil liberties, and human rights. Furthermore, researchers should be integrated who focus on the impact of AI on work, AI governance, bias and discrimination in AI systems, data protection and privacy. Consultation processes that lack broader representation will lack legitimacy and throw into question the authenticity federal government endeavours on AI.
Another angle is addressed in the questions: The issue of research and talent. The question is as follows:
How does Canada retain and grow its AI research edge? What are the promising areas that Canada should lean into, where it can lead the world? (i.e., promising domains for breakthroughs and first-mover advantage; strategic decisions on where to compete, collaborate or defer; balance between fundamental and applied research) Canada should focus on building AI for the common good that puts the needs of workers, communities, and the environment front and centre. The federal government should create a public sector AI innovation hub with researchers, universities, trade unions and government to conduct collaborative research for AI that meets public needs. This research and development should target problems that are identified at the workplace and community level. Canada should invest in research that augments workers, rather than replaces them, centring workers' voices in technology research, development and implementation. Involving workers in defining the problem or opportunity, designing the technical and work process features, educating and training the workforce tends to have better outcomes than more top-down approaches. It increases the likelihood that generative AI tools will be used effectively and improve the quality of workers' jobs. Furthermore, a survey in the United States found that workers who report having an influence on technology in the workplace are twice as likely to report job satisfaction.
You must admit this is interesting. We see that, on the one hand, technological tools can be useful to people, but, on the other hand, they can harm them. I'll continue along these lines:
Canada has an opportunity to lead research on addressing bias and discrimination in AI systems and building targeted AI systems that do not put strain on our energy system and do not drive up greenhouse gas emissions. Canada should invest in AI built on a privacy by design methodology which creates a proactive approach where privacy is embedded as a default setting. The Canadian government should also fund an ecosystem of research into the harms of AI on workers and the public. This research should cover job loss, working conditions, algorithmic management, bias and discrimination, surveillance and monitoring in the workplace and society, privacy and data rights and environmental harms. The research can feed back into a loop to ensure AI development and implementation helps humans, rather than harms them. This moment in time calls for true leadership that puts workers, communities and the environment at the centre of AI research rather than the profit interests of technology corporations.
Another important question is raised:
How can Canada strengthen coordination across academia, industry, government and defence to accelerate impactful AI research? (i.e., mechanisms for cross-sector collaboration; integration of public and private research efforts; industry-sponsored research while preserving academic independence) The Canadian government should establish roundtables where unions, researchers, post-secondary institutions and the federal, provincial and municipal governments meet to discuss AI research. CUPE has called for the federal government to create a Canadian Observatory on AI and Work to develop data and disseminate research on the impact of AI on work. This observatory should include academics, trade unions, civil society and government. As the union-led advisory table report made clear, outside Quebec there are few ongoing consultative bodies where workers, through unions, can systematically gather and share information, coordinate, and negotiate planning. Re-establishing sectoral partnership tables would help identify and address the needs of workers in different sectors and occupations in the context of technological change driven by AI. These tables can develop plans regarding recruitment, retention and succession planning and the employment, training, health and safety and human rights consequences of AI, automation and technological change in the workplace. Multidisciplinary tables should also be established to coordinate research on bias and discrimination and privacy and data rights. The Canadian government should be careful to limit the impact of profit and corporate interests on the research agenda for AI.
I will repeat that: “The Canadian government should be careful to limit the impact of profit and corporate interests on the research agenda for AI.” It is this influence that we are trying to learn more about today by requesting access to the public consultation documents. There may be people who have said things that the government will take into account, but which are not in the public interest. First, we want to know the identity of these people, to determine if they pose a threat to Canada. Second, we want to know if they are acting against the public interest.
In this same paragraph, it states: “Corporations can fund their own research without relying on public funding and resources.” That is the other very important point. As a parliamentarian, I am curious and would like to analyze the content of the comments from the organizations, individuals and companies that participated in the consultation, to determine who is attempting to influence the government to secure public funds for their own interests. However, today, the government is denying us that access. It tells us that the truth comes at too high a cost.
The other question addressed in the brief is: “What conditions are needed to ensure Canadian AI research remains globally competitive and ethically grounded?” Right from the start, we read the following:
(i.e., infrastructure, talent, and governance; ethical standards and risk mitigation; alignment of applied research with business and societal needs) There are a number of enablers to ensure Canada can be at the forefront of research on AI systems for the public good. The federal government will need to commit stable funding and create collaborative spaces and processes for research dissemination. CUPE has proposed that a public sector AI innovation hub be formed with representation from researchers, universities, trade unions and—