Good evening,

In my last AI blog, I wrote about the recent launch of the Canadian AI Safety Institute, and other AISIs around the world. I also mentioned that I was looking forward to learn more about what would be discussed during the International Network for AI Safety meeting that would take place on November 20th-21st.

Well, here’s the gist of it. Representatives from Australia, Canada, the European Commission, France, Japan, Kenya, the Republic of Korea, Singapore, the UK and the US gathered last week in San Francisco to “help drive technical alignment on AI safety research, testing and guidance”. They identified their first four areas of priority:

• Research: We plan, together with the scientific community, to advance research on risks and capabilities of advanced AI systems as well as to share the most relevant results, as appropriate, from research that advances the science of AI safety.
• Testing: We plan to work towards building common best practices for testing advanced AI systems. This work may include conducting joint testing exercises and sharing results from domestic evaluations, as appropriate.
• Guidance: We plan to facilitate shared approaches such as interpreting tests of advanced systems, where appropriate.
• Inclusion: We plan to actively engage countries, partners, and stakeholders in all regions of the world and at all levels of development by sharing information and technical tools in an accessible and collaborative manner, where appropriate. We hope, through these actions, to increase the capacity for a diverse range of actors to participate in the science and practice of AI safety. Through this Network, we are dedicated to collaborating broadly with partners to ensure that safe, secure, and trustworthy AI benefits all of humanity.

Cool. I mean, of course these priority areas are all key to the work that needs to be done… But the network does not provide concrete details on how it actuallyplans to fulfill these priority areas. I guess now we’ll just have to wait and see what actually comes out of it all.

On another note – earlier in the Fall, one of our readers asked us if we had any thoughts about how a win from the Conservatives in the next federal election could impact the future of AI in the country. While I unfortunately do not own a crystal ball, let me share a few preliminary thoughts. 

In May 2024, the House of Commons released the Report of the Standing Committee on Human Resources, Skills and Social Development and the Status of Persons with Disabilities regarding the Implications of Artificial Intelligence Technologies for the Canadian Labour Force.

TL;DR, the recommendations of the Standing Committee notably include: to review federal labour legislation to protect diverse workers’ rights and privacy; to collaborate with provinces, territories and labour representatives to develop a framework to support ethical adoption of AI in workplaces; to invest in AI skills training; to offer financial support to SMEs and non-profits for AI adoption; to investigate ways to utilize AI to increase operational efficiency and productivity; and for Statistics Canada to monitor labour market impacts of AI over time.

Honestly – these are quite respectable recommendations, that could lead to significant improvements around AI implementation if they were to be followed through. 

Going back to the question about the Conservatives, then… The Standing Committee report includes a Dissenting Report from the Conservative Party, which states that the report “does not go sufficiently in depth in how the lack of action concerning these topics [regulations around privacy, the poor state of productivity and innovation and how AI can be used to boost efficiencies, etc.] creates challenges to our ability to manage AI’s impact on the Canadian workforce”. In short, it says do more – without giving any recommendation whatsoever about what that more should be.

On the other side, we know that one of the reasons why Bill C-27 is stagnating is because of oppositions. The Conservatives notably accused the Liberal government of seeking to “censor the Internet” – the Conservatives are opposed to governmental influence (i.e., regulation) on what can or can’t be posted online. But we also know that one significant risk of the rise of AI is the growth of disinformation, deepfakes, and more. So… maybe a certain level of “quality control” or fact-checking would be a good thing? 

All in all, it seems like Conservatives would in theory support a growing use of AI to fight against Canada’s productivity crisis and reduce red tape. In another post previously this year, Alex has also already talked about what a Poilievre Government science policy could look like, and we both agree that the Conservatives at least appear to be committed to investing in technology. However, how they would plan to regulate the tech to ensure ethical use remains to be seen. If you have any more thoughts on that, though, I’d love to hear them. Leave a comment or send me a quick email!

And if you want to continue discussing Canada’s role in the future of AI, make sure to register to HESA’s AI-CADEMY so you do not miss our panel “Canada’s Policy Response to AI”, where we’ll have the pleasure of welcoming Rajan Sawhney, Minister of Advanced Education (Government of Alberta), Mark Schaan, Deputy Secretary to the Cabinet on AI (Government of Canada), and Elissa Strome, Executive Director of the Pan-Canadian AI Strategy (CIFAR), and where we’ll discuss all things along the lines of what should governments’ role be in shaping the development of AI?.

Enjoy the rest of your week-end, all!

– Sandrine Desforges, Research Associate

sdesforges@higheredstrategy.com