Industry and Technology Committee on June 1st, 2026

[Brian Gallant]

That is the real question, isn't it?

I will take this opportunity to say something, not necessarily as the CEO of Space Canada, but as a Canadian. I experiment with language models, just like other people are doing. I am curious about what they know or don't know. I have had the privilege to be featured in the media. I asked a language model a series questions—I knew the answers, of course—and many of the answers I got were often wrong, which left me wondering about the reasons behind that.

For example, I asked the language model what the former premier of New Brunswick does to keep fit. It gave me a series of exercises that I had never seen or done before. I looked up the source, and it turned out to be a satirical site that had published an article about me, suggesting that that was my exercise routine. However, the language model took that as accurate information.

I am extremely concerned about what will happen when it comes to the accuracy of information, but I'm also concerned about how people will be influenced and the decisions that will be made using artificial intelligence when those decisions are based on sources that should never be considered.

[Karim Bardeesy, Liberal, Taiaiako'n—Parkdale—High Park, ON]

Thank you.

Mr. Gallant, you've talked about ambition. SpaceX, one of the largest private companies, is preparing to go public and has linked this to deploying artificial intelligence in space.

What do you, and your members, think about this? What should Canadians think about this type of ambitious plan?

[Brian Gallant]

My colleague has already spoken to the challenges associated with monolithic foreign businesses or the ones that have so much power that they can disrupt the operations of a company based in another country or curb a country's ambitions. That is very concerning. This was not a major concern two or three years ago, but things have now changed, obviously.

I'll say this just to ensure that I can get the right quote. There is often talk of space defence—I'm assuming AI will be similar moving forward—where the government plays an important role in making investments. I'm just going to take the opportunity to mention SpaceX, a company that so many people see as incredibly innovative.

There are obviously arguments for that to be the case, but I'll quote a Harvard Business Review article from a year or two years ago—

[The Vice-Chair, Conservative, Raquel Dancho]

I'm so sorry. We're almost a minute over time.

[Brian Gallant]

I missed the punchline. That's all right.

[The Vice-Chair, Conservative, Raquel Dancho]

I thought we could come right back, Mr. Bardeesy, if that works for you. Okay.

Mr. Ste-Marie, you have two and a half minutes, please.

[Gabriel Ste-Marie, Bloc, Joliette—Manawan, QC]

Thank you, Madam Chair.

First, I want to apologize because earlier, I thought you had a six-minute round. You did a better job managing your time than I did, and so I'd like to apologize.

Mr. Gallant, you can finish what you were saying if you'd like to.

[Brian Gallant]

Thank you, that is extremely kind of you.

I just wanted to add that the Harvard Business Review had an article about two years ago. It said, “Roughly 90% of the first $1 billion invested in SpaceX came from NASA's contracting arrangements, bringing some predictability to an inherently risky venture.”

That figure shows how much strategic industries need governments to invest in the capabilities that we need.

[Gabriel Ste-Marie, Bloc, Joliette—Manawan, QC]

Thank you very much for wrapping up.

Alain McKenna the journalist has written an article on artificial intelligence in today's issue of La Presse. He notes that “Because of cost, companies that rushed to deploy artificial intelligence, such as Microsoft, Uber and Nvidia, are cutting their AI budgets and are even questioning the relevance of massively deploying AI across the board”.

Finally, companies are going to make better use of their resources for cutting-edge developments, but they are either going to revert to using employees for more basic tasks or to using older artificial intelligence models that consume fewer tokens.

Have you heard about this trend from your members?

[Brian Gallant]

I've not seen the article, but there is no doubt that some of these trends are very important and provide a sense of how things are going. I think we might end up with some companies having a lot of power and influence.

I know that this is complex and that it's not quite the same thing, but we allowed it to happen with social media. A few companies have a lot of power and influence due to a lack of regulations or a lack of strategy to develop regulations in other countries. Furthermore, obviously, there are trade tensions related to the non-taxable revenues in countries where these companies do business. To me, that is concerning.

Additionally, if Canada wants to have a system where a few other players try to develop artificial intelligence, will it be as advanced as some of the other countries? That's another thing that concerns me a lot, particularly when it comes to smaller organizations.

[The Vice-Chair, Conservative, Raquel Dancho]

Thank you very much.

We are technically out of time for this hour.

Colleagues, if you want to do the last five-minute round, we might continue. I just want to consult you to see whether that works for everyone. It will bump the rest of our afternoon, so I want to make sure it's all right.

Everyone's fine with it. That's good.

Ms. Borrelli, you have five minutes.

[Kathy Borrelli, Conservative, Windsor—Tecumseh—Lakeshore, ON]

Thank you, Chair.

Welcome to all witnesses. Thank you so much for being here today to answer our many questions.

My question is for Mr. Gallant.

You said earlier today that AI adoption has been very slow in Canada. I recently read that we've invested less than one one-thousandth of what has been invested in the United States.

To date, has Canada invested enough funds to encourage the development and use of AI?

[Brian Gallant]

My humble opinion is no. This goes to what I was saying earlier. We have to make some big plays in some of these sectors.

If I may, I'll make an analogy with the space sector. As a global sector, it's projected to grow from $600 billion in 2021 dollars, give or take, to $2 trillion by 2040. The government, I have to say, has made some big plays for space over the last few months, and it's greatly appreciated. However, in the years we weren't doing that, many others were making big plays.

I would certainly make that analogy. For space, cyber, AI and quantum, we really have to double down, and not just because there's immense potential, economically speaking. If we don't adopt those technologies, our productivity and competitiveness are going to suffer greatly.

[Kathy Borrelli, Conservative, Windsor—Tecumseh—Lakeshore, ON]

Thank you for that.

Mr. Gallant, again, a lot of our AI discussion has been focused on data centres and cloud computing. We haven't focused on space systems that generate the data that AI tools increasingly depend on.

Should Canada treat satellite earth observation and launch capacity as integral parts of the national AI infrastructure?

[Brian Gallant]

To the extent that there will be data, it absolutely should.

As I mentioned earlier, and as my colleague mentioned, there are differences among the types of AI that would be deployed. It is really important for us to make those distinctions when developing any sorts of rules or frameworks around this. Absolutely, it's going to be important for certainty, as well—for companies to understand what the rules of engagement will be.

I can't stress this enough. It's the second or third time I'm saying it, but I'll take the opportunity. One thing I certainly encourage this group to push the government to do is to consistently consult on AI, because things are going to change and be more complicated. We will have to be very nimble and, I would argue, action-oriented if we're going to mitigate those challenges and seize the opportunities.

[Kathy Borrelli, Conservative, Windsor—Tecumseh—Lakeshore, ON]

Space Canada has recommended that there be a national space council chaired by the Prime Minister.

What problem would this solve that the current departmental structure cannot, and what exactly would that council be looking at doing?

[Brian Gallant]

Thank you very much for the question. Maybe I'll go to the challenge, and then quickly say why this might be a good solution.

A challenge for a sector like space is that it is emerging and has changed. Honestly, I think it's very similar to what we're all seeing with AI. To be nimble and able to make the proper investments and policies we need in order to seize the economic opportunity represented by space, we need departments working together and working differently when it comes to space.

We proposed, as a solution, a national space council that would bring in the departments and get them working together. Since we made that recommendation, there has been some moves by the government to try to bring down silos, which we really appreciate. We still argue that there's a need for some type of body to ensure that we're as nimble as possible, as a Canadian space ecosystem.

Frankly, the same would probably apply to AI. I know there are discussions to ensure that there are types of advisory boards and councils to help do that.

[Kathy Borrelli, Conservative, Windsor—Tecumseh—Lakeshore, ON]

Canadian space and AI companies often struggle to find domestic customers. Should federal procurement be used more deliberately to give Canadian companies first opportunities in strategic technologies? How can we do that?

[Brian Gallant]

Yes. That's the first part of the answer.

My members will often tell me and remind government officials that what is helpful to them is that they have anchor customers and earlier customers, as you alluded to in your question, and how it helps to have government.... Not only does it give cash flow for some of the start-ups and not only does it give them the opportunity to know there are some investments coming down the road, but when they go internationally, they can say, “The Government of Canada is a customer of ours. We would love for you to be a customer.” That certainly resonates a lot more than saying our own government isn't supporting us.

The only other thing I would add to the way you asked the question, which is very good—

[The Vice-Chair, Conservative, Raquel Dancho]

Just briefly, please....

[Brian Gallant]

Well, I won't add it, then.

[Voices]

Oh, oh!

[The Vice-Chair, Conservative, Raquel Dancho]

Okay. I'm so sorry.

This is a tougher job than it looks, colleagues.

Ms. O'Rourke, you have five minutes, please.

[Dominique O'Rourke, Liberal, Guelph, ON]

Thank you, Chair.

My questions are for Mr. Reid and Mr. Gallant.

Mr. Reid, thank you so much for mapping out that we're conflating all that's AI into the same conversation, and I think we're having trouble finding our way out. You're talking about how large language models are different from an application with a closed dataset, which we're seeing applied in a number of places, and how then, also, far off in the future, there is this sort of superintelligence. We're not there yet—is that right? Thank you for that.

You also talked about full stack. The Munk School has “Sovereign by Design”, and they've created a heat map that shows where Canadian AI sovereignty is critical. In some of the applications, if you want to make a funny photo or something, that's not a high data-sovereignty place, but space likely will be.

I'm wondering if you are familiar with this model, which looks at where the vulnerabilities are in the system and focuses on those, and if you agree with that system, and then if you would put space and the data gleaned from space as part of that critical infrastructure.

If you can answer quickly, I have another couple of questions.

[Ewan Reid]

I'm not familiar with the model in that specific example, but I absolutely agree that space would be absolutely critical. I'm sure our Canadian Armed Forces would agree with that.