Information & Ethics Committee on Feb. 2nd, 2026

[Michael Barrett, Conservative, Leeds—Grenville—Thousand Islands—Rideau Lakes, ON]

First of all, based on their co-operation to this point, is it your expectation they'll meet the deadline? Should they not, what would be the outcome?

[Philippe Dufresne]

Should they not, one of the gaps in privacy legislation is that I don't have order-making powers. We have to count on their collaboration and we have to work with them. So far, it is working and I'm cautiously optimistic that it's going to continue to work, but that is why a key element of law reform has to be order-making powers for the Privacy Commissioner.

[Michael Barrett, Conservative, Leeds—Grenville—Thousand Islands—Rideau Lakes, ON]

I'm pressed for time, so I'm going to move on.

I want to ask about your investigation of X Corp. This is incredibly important, obviously, the harms being done in real time, not only to children but especially to children, with the non-consensual images that are being propagated and created are alarming.

What kind of a position are you in, in your role as a Canadian commissioner, with the lack of order-making power? Do you lack the enforcement tools necessary to be able to effectively execute your role in a way that would protect Canadians, especially children, in a situation where the gravity of it can be lost on no one?

[Philippe Dufresne]

I'll say two things.

One is that I need to have order-making powers. That should be a top priority for Parliament. I'm encouraged by statements from the Minister of AI that modernization is something that's being contemplated.

That being said, I am using and will continue to use the existing tools that I currently have to protect Canadians' privacy with investigations and recommendations and by working with partners in Canada and internationally.

In the context of deepfakes, we issued statements calling them out as early as December 2023. We're working with international partners to bring the international community together in taking steps.

That's one thing that I've done, for instance, in the 23andMe case with the breach of Canadians' information. I don't have order-making powers but my U.K. counterpart does, and together, we did that for Canadians.

[The Chair, Conservative, John Brassard]

Thank you, Mr. Dufresne and Mr. Barrett.

Mr. Sari, you have the floor for six minutes.

[Abdelhaq Sari, Liberal, Bourassa, QC]

Thank you, Mr. Chair.

Thank you very much for your testimony, Mr. Dufresne.

First, before I ask my question, I’d like us to talk about your mandate regarding privacy in general. The mandate is based on fundamental principles, including the issue of transparency, general accountability and the ability for Canadians to understand and even challenge the decisions made by any means, whether powered by artificial intelligence or any other solution.

However, as you well know, artificial intelligence has somewhat changed how decisions are made. For example, all artificial intelligence, whether generative, cognitive or agentic, ultimately relies on neural networks, which don’t provide a decision process. This is what is called the black box. There is no model on the decision-making process.

My first question is as follows:

In this kind of scenario, where technologies are evolving at a very fast pace, how is the Office of the Privacy Commissioner evolving? How does it assess the existence of actual human oversight? Can we talk about a certain degree of human control?

[Philippe Dufresne]

Thank you for your question.

You’re quite right. Privacy principles can evolve over time because these are fundamental principles that affect human dignity, freedom, the ability to make decisions, the ability to choose what we share and with whom. Right now, we’re facing serious challenges with technology, because technology is evolving very quickly. Organizations like mine generally need to catch up. They need to learn about these tools and to develop technological capabilities. My office has developed capabilities to understand the technology and to act accordingly.

On the issue of human decisions that you alluded to, we also need to raise these issues in the course of legislative reforms. For instance, I made a recommendation calling for more transparency with artificial intelligence to understand the decisions that are made and to make sure humans are in the process. Recently, I was in Seoul for the global privacy assembly, and my office sponsored a resolution, which was adopted by the international community, that focused on the human role in the artificial intelligence process, its significance and what it must embody. This issue is therefore timely, and it’s somewhat of a response to the black box effect you mentioned.

[Abdelhaq Sari, Liberal, Bourassa, QC]

The black box is more technological. It’s the way neural networks generally operate, and there is no way to walk that back and change it. It started in the 1980s and it will continue to work that way.

You talked about understanding technology. Do you think the right to privacy is still being fully respected?

[Philippe Dufresne]

I think there are definitely cases where it’s not respected. We talked about the TikTok decision, where there were shortcomings. There is also the recent decision concerning Staples. There are many cases where there have been shortcomings, but the key thing is to have the tools to address these shortcomings. Ideally, we take preventive measures, but if not, we take action after the fact.

I believe these principles are technologically neutral and they can evolve with technology. However, the law needs to be modernized to make this easier. Concepts like de-identification, transparency and how to process transparency when organizations themselves say they don’t know how they made such decisions. The chair of Google said that recently. The law must therefore change. We can have more proactive obligations, but we already have a very good framework on this issue with the Privacy Act.

[Abdelhaq Sari, Liberal, Bourassa, QC]

It makes me feel better to hear you talk about the framework because all the scientific publications that I have read, and the people who have appeared before this committee, put more emphasis on the lack of transparency with artificial intelligence technologies. I understand and have a lot of respect for your role, which I think is very important. You have reassured me a lot today because of the progress you have made in your work.

On our part, as a House of Commons committee, what would you recommend that we do with respect to oversight? What steps must we take to manage this transparency gap in oversight, recognizing that some lack of transparency is inevitable, whether we like it or not?

[Philippe Dufresne]

In my opinion, the most important thing you can do as parliamentarians is to amend the law. It’s an essential tool. I made seven priority recommendations to the government, which I shared with the committee. They include such items as expanded enforcement-making powers, order-making powers and the power to impose fines in certain cases. Recognizing privacy as a fundamental right is important. This is more important than ever with the advent of artificial intelligence. The recommendations include items such as privacy by design. Organizations that create tools that have significant impacts must be asked to conduct privacy impact assessments right from the beginning, and not at the end. This would be positive across the board. This issue must also be addressed globally. We need a framework to govern data that leaves Canadian jurisdiction.

[Abdelhaq Sari, Liberal, Bourassa, QC]

In terms of information, I always say there’s a need to collect, process, store, secure and share information globally using other factors. Your role is particularly meaningful when it comes to security and protection.

Let us turn to Canadians listening to us today. Is there one group that tends to be more vulnerable than others? Who is the most vulnerable? We have talked a lot about young people and women today. Are certain groups of people more vulnerable than others?

[The Chair, Conservative, John Brassard]

Please give a very brief answer, Mr. Dufresne.

[Philippe Dufresne]

Young people are very vulnerable because they are immersed in it, they experience it from every angle. You’re right to allude to women and seniors. I believe many groups need to be protected.

[The Chair, Conservative, John Brassard]

Thank you, Mr. Sari.

Mr. Thériault, you have the floor for six minutes.

[Luc Thériault, Bloc, Montcalm, QC]

Thank you, Mr. Chair.

Commissioner, welcome.

You talked about the international aspect earlier. I wanted to bring you back to—

[Abdelhaq Sari, Liberal, Bourassa, QC]

Mr. Chair, I’m having trouble hearing Mr. Thériault.

[The Chair, Conservative, John Brassard]

Just a moment.

Please lower your mike, Mr. Thériault. That’s better. Thank you.

I will restart the clock.

[Luc Thériault, Bloc, Montcalm, QC]

Okay, thank you.

I wanted to take you to the G7 leaders’ statement, which reads as follows:

Work together to accelerate adoption of AI in the public sector to enhance the quality of public services for both citizens and businesses and increase government efficiency while respecting human rights and privacy, as well as promoting transparency, fairness, and accountability.

Further down, it states:

Promote economic prosperity by supporting SMEs to adopt and develop AI that respects personal data and intellectual property rights, and strengthen their readiness, efficiency, productivity and competitiveness.

In your opinion, for G7 leaders, what should be the key elements of a human-focused approach to artificial intelligence?

[Philippe Dufresne]

Thank you for your question.

You’ve alluded to the G7 leaders’ statement. The G7 privacy commissioners, including me, had our own meeting here, at Meech Lake. In June, we issued a unanimous statement on the importance of prioritizing privacy for two reasons: to promote innovation and the economy and to protect children and young people.

We alluded to the leaders’ statements on these items, and provided concrete examples including considering the best interests of children and young people when technologies are designed, developed and brought to market; underscoring the importance of privacy at the beginning, and not at the end; and underscoring the importance of transparency and communication with Canadians to create trust.

That is why some people tend to say that a strong privacy regime can stifle innovation. I challenge that, and our G7 statement also challenged that. The statement clearly says that it will promote innovation and support economic development because it will build public and consumer trust in this new technology. This can be seen in the surveys I alluded to earlier. There are concerns on this issue. It’s therefore essential from a fundamental rights standpoint if we want to protect children, women, seniors and the rest of us when it comes to our freedom and our dignity. Furthermore, pragmatically, it will support the global economy and Canada’s economy in particular. We sent a strong message here in Ottawa. I was very proud of that statement, and we are still promoting it here and beyond.

[Luc Thériault, Bloc, Montcalm, QC]

When it comes to enhancing the quality of public services and increasing government efficiency, what is the best way to balance the speedy adoption of artificial intelligence with respect for human rights and privacy, and advancing transparency, fairness and accountability?

[Philippe Dufresne]

That’s the point. It’s important to strike that balance. Tools will increase productivity. They can support all sorts of economic development. They can support developments in health care. They can improve services for citizens. That’s very positive. However, this must be done in a transparent and humane manner. The concept must be human-centred. It must be done in a manner that respects the fundamental principles of privacy, such as informed consent, in certain cases, or appropriate goals.

In September 2025, the office of the Privacy Commissioner of Canada, together with counterparts from the Competition Bureau, the Canadian Radio-television and Telecommunications Commission or CRTC, and the Copyright Board issued an analysis on synthetic media, commonly known as deepfakes, which addressed these issues. In some cases, technology can help to create images. However, if it’s used to manipulate people or if it’s used to create non-consensual sexual images, then, of course, that infringes upon rights that must be respected. We cannot place innovation on one side and privacy on the other. Canadians must have both.

[Luc Thériault, Bloc, Montcalm, QC]

Take work, for example. It’s a fundamental right. We can say it’s part of human rights. People have come to us and said that states will want to replace white-collar workers with artificial intelligence soon. When a critical mass of people lose their jobs, how can this be reconciled with advances in artificial intelligence? We have talked about the G7 here, and its members have ordered governments to be productive and efficient.

[Philippe Dufresne]

I think the issue of job protection is slightly outside the scope of my privacy mandate. However, I can tell you that when you protect privacy, you protect individuals. You make things more transparent, and you provide a better understanding of events. You also help people to distinguish between what is human and what is not. I believe that can be part of both the goals and the solutions. When it comes to jobs, we can say that we don't want to compete with artificial intelligence where it's a pure case of statistical probability, but the human advantage needs to be retained. In my opinion, that will always be an asset.

[The Chair, Conservative, John Brassard]

Thank you, Mr. Dufresne and Mr. Thériault.

We will now begin the second round of questions. Mr. Cooper, you have five minutes.

Please go ahead, Mr. Cooper.