Evidence of meeting #129 for Access to Information, Privacy and Ethics in the 44th Parliament, 1st Session. (The original version is on Parliament’s site, as are the minutes.) The winning word was research.

A video is available from Parliament.

On the agenda

MPs speaking

Also speaking

Mireille Lalancette  Professor, Political Communication, Université du Québec à Trois-Rivières, As an Individual
Timothy Caulfield  Professor, Faculty of Law and School of Public Health, University of Alberta, As an Individual
Marcus Kolga  Director, DisinfoWatch
Yoshua Bengio  Founder and Scientific Director, Mila - Quebec Artificial Intelligence Institute

Matthew Green NDP Hamilton Centre, ON

I hear often that we have a digital axis of evil. We talk about China. We talk about Russia. We sometimes talk about India.

Is it not safe to say that all countries are engaging in this in some form or another and that it's often a matter of degrees of rhetoric about what is an influence campaign and having international influence, versus international interference?

Going back to this idea of treaties and of internationally recognized law being established, you've mentioned the United States and Europe. Are they not also involved in this? If so, are the regulations they're putting forward only considering the short-term and domestic interests?

5:20 p.m.

Founder and Scientific Director, Mila - Quebec Artificial Intelligence Institute

Yoshua Bengio

I think there are ethical red lines. We do have legal protections that don't exist in other countries, so there are differences. You're right. Everyone's trying to use technology to their advantage in different degrees.

We should make sure that we have more transparency in how these tools are used, whether it is by corporations or by state actors. Our governments eventually might be tempted to use AI in order to influence their own people. We need to have guardrails against that as well. Of course, we need to protect against state actors who are clearly intent on destabilizing our democracies.

You have to think that in a few years from now—maybe a decade, I don't know—we're going to build machines that are going to be as smart as humans. At least it's very plausible, and the majority of AI researchers think we're going to get there. How is that going to be used? There's a chance that there's going to be an abuse of that power by whoever controls these machines.

5:20 p.m.

Conservative

The Chair Conservative John Brassard

Thank you.

Matthew Green NDP Hamilton Centre, ON

Thank you, Mr. Chair, for the indulgence.

5:20 p.m.

Conservative

The Chair Conservative John Brassard

Yes. No worries.

We are going to the second round now, starting with Mr. Cooper for five minutes.

Go ahead, Mr. Cooper.

5:20 p.m.

Conservative

Michael Cooper Conservative St. Albert—Edmonton, AB

Thank you very much, Mr. Chair.

Mr. Kolga, when you appeared before the procedure and House affairs committee back in November of 2022 as part of its study on foreign election interference, you gave testimony. I asked you some questions about Beijing's disinformation campaign in Steveston—Richmond East. Since then, a lot of new information has emerged from the public inquiry and in other fora, including this committee.

Last week, the office of the Commissioner of Canada Elections tabled a report at the public inquiry into foreign interference in federal electoral processes and democratic institutions. The report confirmed that the PRC targeted former MP Kenny Chiu in Steveston—Richmond East as part of a disinformation campaign designed to drive voters away from him, using fairly sophisticated efforts, including through the amplification of such disinformation on social media platforms.

Madam Justice Hogue, in her initial report back in the spring, noted that, “there is a reasonable possibility that these narratives could have impacted the result in this riding.” In other words, the disinformation was significant enough that it could have tilted the balance in terms of the outcome of the election in that particular riding, resulting in the defeat of Mr. Chiu.

I find that alarming. Do you? It seems to me that it should sound the alarm on how serious the threat of foreign interference from hostile foreign states, including the PRC, is to our democracy and the integrity of our elections.

5:25 p.m.

Director, DisinfoWatch

Marcus Kolga

I have no doubt that that operation did impact the result.

As to whether it tilted the result one way or the other in any significant way, we don't know that. There's no doubt that it had some form of impact.

What to do about this in the future? Again, I will go back to my statement earlier. We tried to raise the alarm about that operation as it was happening or shortly after we observed it happening. This is something that needs to be done in the future.

I'm not sure that it's the government necessarily that has a role in doing that. Certainly we should be empowering civil society organizations, those that are monitoring the information space, who understand it and who understand where to look: Empower those groups and individuals who are doing that work so that they can alert Canadians and the Canadian media to those sorts of efforts when they're happening. That will help to build resilience and strengthen our democracy.

5:25 p.m.

Conservative

Michael Cooper Conservative St. Albert—Edmonton, AB

Well, you indicated that perhaps it's not the role of government, but there were structures and processes in place at the time of the 2021 election.

I would note that in May of 2021, an IMU, an issues management note, was sent to the Minister of Public Safety that specifically noted that Kenny Chiu's riding was of high interest to the PRC. That IMU went into a black hole. The minister claims he never saw it, even though it went to him, Minister Blair, his chief of staff and the deputy minister of public safety.

Following the Global Times article of September 9, 2021, the rapid response mechanism at Global Affairs Canada highlighted the disinformation that was being spread. That was provided to the election committee that had been set up by the government to counter foreign interference and to bring awareness to interference activities.

Through it all, Kenny Chiu was left to drown in a sea of disinformation. Not only did he drown; in fact, there is evidence that the Liberal Party actually amplified the disinformation being spread by the Beijing-based regime. Not only is it the case that the PRC intervened in the Steveston—Richmond East riding; one of the major political parties, namely the Liberal Party, actually amplified the disinformation, perhaps resulting in the defeat of Kenny Chiu.

Would you agree that the system failed Kenny Chiu and failed the voters of Steveston—Richmond East?

5:25 p.m.

Conservative

The Chair Conservative John Brassard

I'll have to ask you to do this very, very quickly, Mr. Kolga, please.

5:25 p.m.

Director, DisinfoWatch

Marcus Kolga

I would absolutely say that there's room for improvement within that system to ensure that when these instances are detected in upcoming elections, they are brought to the attention of those individuals who are being targeted.

That goes beyond elections; that goes to the spaces between elections. Whether it's elected officials, candidates or activists being targeted by information operations, they need to be notified. There is lots of room for improvement.

5:30 p.m.

Conservative

The Chair Conservative John Brassard

Thank you, sir.

Thank you, Mr. Cooper.

Mr. Bains, you have five minutes. Go ahead, please.

Parm Bains Liberal Steveston—Richmond East, BC

Thank you, Mr. Chair.

Thank you to our witnesses for joining us on this very important study.

I'll draw your attention to Bill C-70. Are you familiar with motion 112, which was also introduced in the House? What we're talking about here is information.

Some important recommendations came out of this study. On the misinformation piece, what Mr. Cooper ended with in his question was that there was misinformation. The Liberal Party did not amplify the claims that he's talking about. We've seen evidence of him drawing different conclusions earlier in this study today as well.

However, I will stick to the important matters here—namely, the solutions and what you're talking about in terms of what we can do better. On the issue of information that's being shared, we can look at the Security of Information Act. In fact, a former director of CSIS came into the study and talked about how important it was to use that as an enforcement tool. I talked about motion 112 and sharing information with hostile nations and some agreements we have that already exist and that need to be reviewed. Motion 112 is something I co-authored with MP Dhaliwal.

Can you expand on how important the Security of Information Act is, within your scope of knowledge on that? What impact can the fact that it hasn't been updated for over 20 years have?

5:30 p.m.

Director, DisinfoWatch

Marcus Kolga

I think it will have a tremendous impact if it's properly implemented, and, again, if it's enforced. That update in Bill C-70 to the Security of Information Act will go a long way to helping defend vulnerable communities, activists, journalists and of course parliamentarians who are being targeted with transnational repression. I think that's positive.

There's the update to the CSIS Act and the fact that the update will allow CSIS to now communicate with non-governmental organizations when it's relevant, perhaps to warn individuals and groups when they are being targeted by these sorts of operations and to let civil society organizations know when they detect these sorts of operations, so that we can be better prepared to defend Canadian democracy and society against them. That's a major improvement.

As part of Bill C-70, we also saw the adoption of the foreign influence transparency registry, which will also be critically important. The implementation we'll have to keep an eye on, but this will also be an incredibly important measure in defending against these sorts of operations in the future. I hope that the legislation will be implemented very quickly and that it will be enforced.

Parm Bains Liberal Steveston—Richmond East, BC

I'm going to ask a different question and take a different path.

Do you think there needs to be more vigorous background checking of people who are interested in running for office and who may have had previous relationships with foreign entities and foreign countries?

For example, in 2020, the current member for Calgary Heritage helped produce a controversial report by former CBC reporter Terry Milewski alleging that Pakistan secretly created a Sikh separatist movement. This report was then amplified by official Indian government accounts. Later, the same member led the Macdonald-Laurier Institute's partnership with the New Delhi-based Observer Research Foundation, an Indian think tank set up with funds from Indian oil giants, and is now a member.

Do we need to look at some relationships members like this may have had with previous governments that are now showing hostilities and interfering domestically?

5:30 p.m.

Director, DisinfoWatch

Marcus Kolga

Absolutely. That process of vetting any candidates at all three levels of government should be far more rigorous. I don't think anyone would have anything against greater transparency as part of that process. I would completely support that.

Parm Bains Liberal Steveston—Richmond East, BC

How much time do I have, Mr. Chair?

5:35 p.m.

Conservative

The Chair Conservative John Brassard

You have five seconds, Mr. Bains.

Parm Bains Liberal Steveston—Richmond East, BC

Thank you for your testimony today. It's very valuable.

5:35 p.m.

Conservative

The Chair Conservative John Brassard

Thank you, Mr. Bains.

Thank you, Mr. Kolga.

Mr. Villemure, you have the floor for two and a half minutes.

René Villemure Bloc Trois-Rivières, QC

Thank you, Mr. Chair. I'll try to make good use of these two and a half minutes.

Mr. Bengio, I'll turn to you again.

Could you, for a moment, take a forward-looking view, describe the future of artificial intelligence, and if possible, give us some examples?

5:35 p.m.

Founder and Scientific Director, Mila - Quebec Artificial Intelligence Institute

Yoshua Bengio

I'd like to thank you for asking me this question, because I don't think we look far enough into the future.

We have to understand that human intelligence knows no absolute limits. It's almost certain that we'll be able to build machines that will surpass us in many areas. We can't know for sure whether this will be in a few years or a few decades, but we need to be prepared for it.

What I find perhaps most worrying is that this means that those who will control these systems will have immense power, whether they be states, companies or others. I mention this because today we're concerned with protecting democracy. We're going to have to set up safeguards to make sure we don't have too much power concentrated in one place, whether it's in the hands of one person, a company director, any other organization, or even a government. The greater the possibilities of these systems become, the more important the question of governance will become.

It's a bit like creating entities or a new species whose intelligence might surpass our own. It's a very dangerous thing. We need to exercise control over this to ensure that artificial intelligence remains a tool, and not something that could compete with humans. We're talking about something much further away in time, but people at companies like OpenAI and Anthropic think it could happen as quickly as five years from now. So we need to start worrying about it today.

René Villemure Bloc Trois-Rivières, QC

We indeed have a legislative role here today. What can we do right now, in addition to Bill C‑27, so as to understand and meet these challenges of the future that are rushing towards us?

5:35 p.m.

Founder and Scientific Director, Mila - Quebec Artificial Intelligence Institute

Yoshua Bengio

I think we need to give incentives to companies to develop better protections. The government needs to fund research so that artificial intelligence systems tend towards public protection. Earlier, we were talking about fact checking, for example. We need to design tools that will help our intelligence and national security communities protect themselves against attacks that could come from other countries using artificial intelligence. With the help of our international partners, we need to develop ways of making artificial intelligence secure.

However, all these issues are more a matter for governments than for the companies that make gadgets. The latter want to maximize their profits. They're not in it for the collective good. They compete with each other to sell us as many things as possible, which is perfectly normal in our market system. However, this means that the responsibility lies with governments. They're the ones who have to deal with it.

René Villemure Bloc Trois-Rivières, QC

Thank you very much, Mr. Bengio.

5:35 p.m.

Conservative

The Chair Conservative John Brassard

Thank you, Mr. Bengio and Mr. Villemure.

Mr. Green, you have two and a half minutes. Go ahead, please.