Evidence of meeting #25 for Science and Research in the 45th Parliament, 1st session. (The original version is on Parliament’s site, as are the minutes.) The winning word was systems.

A recording is available from Parliament.

On the agenda

Members speaking

Before the committee

Mehmet Murat Kristal  Professor, Schulich School of Business, York University, As an Individual
Taylor Owen  Beaverbrook Chair in Media, Ethics and Communications and Founding Director of the Centre for Media, Technology and Democracy, As an Individual
Steven Murphy  President and Vice-Chancellor, Ontario Tech University
Peter Lewis  Canada Research Chair in Trustworthy Artificial Intelligence, Ontario Tech University
Hinton  Intellectual Property Lawyer, As an Individual
Nguyen  Chief AI Officer, Conseil de l'innovation du Québec
Tijs Creutzberg  President and Chief Executive Officer, Council of Canadian Academies

12:25 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

Thank you for your question.

I'm actually not well prepared to speak to that question. This is not something we have looked at in the CCA or with any of our panels.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

I was wondering if perhaps you could speak to AI deepfakes being used in scams. Is that something that you've researched?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

We have not directly, no. We have looked at online harms and the like, but not specifically....

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

Maybe in a broad sense, what does Canada need to do to develop AI responsibly and be a trusted international partner?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

That's a good question.

Once again, this is not something that we've explicitly taken on, but broadly.... I'll leave it there. These are good questions to be putting to us, but I don't have a ready answer for you on that.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

Perhaps you could give an overview of what you've been researching lately.

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

Certainly. We've been looking in depth at the state of our science, technology and innovation ecosystem, including looking at AI. We're looking at how AI is being diffused through the economy.

We've looked a bit at AI research here in Canada. We've looked at the security environment of research. As I mentioned, that's the tension between open science and research areas of concern, which include AI. We're now looking at geodetic infrastructure and at some of the dependencies we have on U.S. infrastructure.

Similarly, in the area of meteorological services, we've been looking at our system and some of those dependencies on the U.S.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

Could you elaborate on the security that you mentioned and the sovereignty of the AI infrastructure here in Canada? Are there any gaps that you see right now that we need to address?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

On the AI security, apart from some dependence on U.S. infrastructure, in the context of research there are gaps. The big issue here within the research community is having to learn how to deal with this tension between open science and research security. It's a big mind shift for the academic community.

We talked about Huawei. The universities are still reaching out to Huawei and still engaged in partnerships. This takes time. The research community needs to be learning about this and changing their mindset as to how they think about their research and the impacts they're having in the broader geopolitical economy. It's something that they have never had to do in the past. It does take time and that's why we're seeing some lags, I think.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

In terms of infrastructure that we can invest in as a government, what is the most important thing that you see right now? Is it data centres? Is it something else?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

That's a good question. Is it specifically in terms of AI sovereignty? Yes.

I don't know if I can give you a decent answer on that. Again, making recommendations on the infrastructure is not something we've looked at explicitly, so I'll pass on that.

Thanks.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

No worries.

Could you speak to anything we can do to keep IP within Canada? How do we attract researchers in the AI sector to Canada?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

The government has certainly made a move to bring talent into Canada. It's not clear to me at this stage how much of that is focused on AI.

In terms of keeping patents in Canada, we have to improve our innovation ecosystem here. The panel is very clear that the need of Canadian start-ups, whether they're in AI or in other areas, to go to the U.S. for capital means that we lose the IP in the process. It transitions out of the country. We see a net IP outflow going out of the country as a result of those financing arrangements.

Fixing the capital environment for our small and medium-sized enterprises, which are doing a lot of research at the moment, is a key step to help fix the outflow of IP.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

To dive in a bit further, what can we do to make things more competitive here? Is it higher salaries for researchers? Is it tax credits?

12:30 p.m.

President and Chief Executive Officer, Council of Canadian Academies

Dr. Tijs Creutzberg

There are a lot of interdependent pieces to this problem. Without retaining those larger companies in Canada, we don't have the opportunities, and without the talent, we're not able to build out those companies.

There are a lot of pieces that need to come together in order to make the shift. It's not one thing; it's several things. I think it needs a concerted effort and a commitment in the longer term to pull it off.

Jake Sawatzky Liberal New Westminster—Burnaby—Maillardville, BC

Thank you very much.

The Chair Liberal Salma Zahid

Thank you. Your time is up.

We will now go to MP Blanchette-Joncas for six minutes. Please go ahead.

Maxime Blanchette-Joncas Bloc Rimouski—La Matapédia, QC

Thank you, Madam Chair.

I'd like to welcome the witnesses who are with us today.

My first question is for you, Mr. Hinton.

Is there a binding legal mechanism to stop strategic AI assets that are, for example, publicly funded from leaving the country?

12:35 p.m.

Intellectual Property Lawyer, As an Individual

Jim Hinton

No. There are announcements about intentions to put pieces in place, but ultimately we see Canada's AI assets fleeing the country at an alarming pace. Our data is out the door very quickly. Our algorithms and our patents...all of that is there. There's no strategic oversight. There are no mechanisms. We, in fact, have mechanisms to accelerate the ability of Canada's AI technologies to leave the country.

Maxime Blanchette-Joncas Bloc Rimouski—La Matapédia, QC

Public subsidies, or taxpayer dollars, Canadians' money collected through taxes, can be used, particularly in AI, to consolidate the intellectual property of potential foreign actors. Is that right?

12:35 p.m.

Intellectual Property Lawyer, As an Individual

Jim Hinton

In Canada, we fund our research institutions to exfiltrate value around artificial intelligence to be owned by foreign firms. We know Geoff Hinton—no relation—is a great researcher, but his research is owned by Google technologies. The chair of Google, as he was at the time, said, “Thank you, Canada, for all of the foundational AI research.”

Do you know how much we get? We get nothing. Ozempic and all of these other drugs come out of the University of Toronto, and we get nothing. It's a trend that we have. We create a lot of great stuff and then we incentivize all of that great stuff to be owned and capitalized. There's Novo Nordisk. Those jurisdictions get the economic rent. We don't.

Maxime Blanchette-Joncas Bloc Rimouski—La Matapédia, QC

I'd like to take advantage of your presence here today and your expertise as a lawyer to ask you this: In your opinion, is there any independent authority in Canada today with legal power that could immediately disable a dangerous AI system?

12:35 p.m.

Intellectual Property Lawyer, As an Individual

Jim Hinton

No, there is nothing. It's the wild west. It's anarchy.

Our digital systems do not have any guardrails. There's the Privacy Act from 25 years ago. There are older things from before my time that still permeate, but there is nothing to stop our digital systems and AI systems from doing whatever they do, and that's by design. American firms love the lack of guardrails. That's a policy choice, because it allows them to extract more rent from the Canadian economy—the American economy as well—and other jurisdictions.

We're implicitly supporting the extractive nature of the data-driven economy by having no guardrails and no mechanisms to take harmful AI and data-driven systems off-line if they cause harm.

Maxime Blanchette-Joncas Bloc Rimouski—La Matapédia, QC

Let's summarize all that. We're heavily funding AI. The government was very proud to do so in the last budget. They even say they want to be the world leader in artificial intelligence. However, we don't have an independent oversight authority, we don't have any clear deactivation powers, as you've just confirmed, and we lack robust guardrails.

In your opinion, is that a comprehensive strategy or an incomplete strategy? As an expert, how do you see this?

12:35 p.m.

Intellectual Property Lawyer, As an Individual

Jim Hinton

There was never a strategy. It was a research funding approach to say we need to have researchers who work for foreign firms physically located in Canada, but the value and the IP all go somewhere else. It was never a strategy. We asked Innovation, Science and Economic Development Canada for the strategy document. None was ever provided.

People have admitted on the record, in a Globe and Mail article in January 2023, that no strategy ever existed. It was an announcement. It was performative, and it continues to be. The consultation that happened in the fall was performative. It was not serious. I know people on the committee, and they made an effort to feed back into it. However, it was performative, not serious, and we'll have similar outcomes with the next process.