Evidence of meeting #52 for Status of Women in the 45th Parliament, 1st session. (The original version is on Parliament’s site, as are the minutes.) The winning word was young.

A recording is available from Parliament.

On the agenda

Members speaking

Before the committee

Dunn  Associate Professor and Director, Law and Technology Institute, Schulich School of Law, Dalhousie University, As an Individual
Parent  Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne
Tamil Kendall  Director, Partnership for Women's Health Research Canada
Attard-Frost  Assistant Professor, Department of Political Science, University of Alberta, As an Individual
Sestock  National President, Canadian Federation of University Women
Gillespie  Executive Director, Equal Voice Foundation
Ruddick  Coordinator, Election Tracking and Research, Equal Voice Foundation

4:45 p.m.

Conservative

The Chair Conservative Dominique Vien

Good morning, everyone.

Welcome to meeting number 52 of the Standing Committee on the Status of Women. I hope everyone is in good shape for this meeting.

I would like to raise two points before we begin with our guests.

First, if it is okay with you, colleagues, we would take 20 minutes next Wednesday for the instructions on the menopause study. That would help the analysts a bit. We did, in fact, discuss this on Monday. Since it now appears that you are fine with that, we will go ahead. Thank you very much.

Second, in terms of the schedule, I just want to remind you that Monday, November 2, is the deadline to submit your dissenting or supplementary opinions, if any, and Thursday, November 5, is the deadline to submit your list of witnesses for our study on fertility.

Pursuant to Standing Order 108(2) and the motion adopted by the committee on Monday, September 21, the committee is now resuming its study of the impacts of artificial intelligence on women, girls and 2SLGBTQIA+ people.

Today's meeting is taking place in a hybrid format, pursuant to the Standing Orders.

To ensure an orderly meeting, I would like to outline a few rules for witnesses and members.

I would like to remind witnesses that committee members may ask questions in either French or English. If you will need interpretation, please take a moment now to prepare your earpiece and select the listening channel required in order to take full advantage of the time allotted for questions and answers.

Before speaking, please wait until I recognize you by name, if possible. If you are joining via video conference, please click on the microphone icon to turn on your microphone. Please mute your microphone when you are not speaking. For members in the room, if you wish to speak, please raise your hand. For members on Zoom, please use the “raise hand” function.

As a reminder, all comments by members and witnesses should be addressed through the chair. As I often say, I like it when it is fluid. I do not stop conversations unless it is total chaos.

We are going to address sensitive issues, which may affect some people. If that is the case and you feel the need to speak to someone, please do not hesitate to talk to our clerk, Ms. Miller. We will be pleased to assist and support you.

I will now welcome our witnesses.

We have Suzie Dunn, assistant professor, Schulich School of Law, Dalhousie University.

From the Centre pour l’intelligence émotionnelle en ligne, we have Emmanuelle Parent, cofounder and executive director, and from Partnership for Women's Health Research Canada, we have Tamil Kendall, director.

Welcome. We will begin with opening statements.

Ms. Dunn, the floor is yours for five minutes.

Suzie Dunn Associate Professor and Director, Law and Technology Institute, Schulich School of Law, Dalhousie University, As an Individual

Thank you so much.

Thank you for inviting me to speak today on the impact of artificial intelligence on women, girls and 2SLGBTQI+ individuals.

We often hear that women are falling behind on AI use and we're encouraged to embrace it. This is a challenging call as AI does not always benefit women and often puts them at risk. It's impossible to cover the breadth of harms and challenges that AI technology has facilitated and amplified in five minutes, but I'll do my best to highlight some of them.

Risks from AI range from discriminatory datasets that embed sexist, homophobic and transphobic norms into their outputs, repeating and amplifying historic discrimination, to facial recognition technology that has lower accuracy rates on women and racialized faces, putting them at risk of being misidentified for crimes they did not commit or causing challenges with immigration or at borders. When facial recognition technology does work, it's used to surveil and dox women who are targeted by malicious pickup artists and other bad actors who expose their identities online to encourage harassment against them.

AI agents, large language models and natural language processing tools increase the ability of stalkers and abusers to gather information about their targets and amplify the harassment against them. This happens not only in the context of intimate partner violence. It also targets politicians, journalists and human rights defenders. This leads to the silencing of women's and queer voices, who avoid posting information about themselves online to avoid this form of digital harassment.

AI is also used to create and spread discriminatory myths and disinformation that poison digital spaces and makes it unsafe for women and LGBTQI+ people to fully engage in digital life. Additionally, AI is used to facilitate romance and sexual extortion scams. There needs to be a response by government to address these AI-facilitated harms, not just those caused by individuals but also those created and amplified by the companies themselves.

My main area of expertise is on intimate images and non-consensual synthetic intimate images, which most people call deepfakes. It was clear in 2017 that deepfakes were a significant problem for women. Their identities were easily co-opted at this time and their sexual autonomy was removed. The technology to create these images, and the websites dedicated to hosting them, were widely available. One of the most popular sexual deepfakes websites was in fact created and run by a Canadian. It has taken nearly a decade of women being targeted by this harm before the federal government took action on this issue, and there are still significant gaps remaining.

I was recently speaking with a victim of non-consensual sexual deepfakes, who had her images taken from social media and turned into sexual ones—this occurred a few years ago before the criminalization of deepfakes—and there was no remedy for her at the time, despite this being an obvious problem since 2017. The government and courts did not recognize what happened to her as a harm. She was not considered a victim deserving of help, and there were few resources for her to access and a lack of sufficiently resourced organizations who could help her meaningfully. The deep nude apps that likely facilitated the creation of these images are readily online and remain readily online. She asked me at the time why the government took so long to take these issues seriously and tragically wondered whether it would take someone committing suicide for these issues to be paid attention to seriously.

We shouldn't have to wait for tragedies like the deaths of Amanda Todd, Rehtaeh Parsons or the children at Tumbler Ridge in order for the government to take action on technology-facilitated harms. Canada needs nationally available resources and supports for AI-based, technology-facilitated, gender-based violence, as well as victim-centred legal responses.

We've seen some real advances in provinces like British Columbia on issues like intimate images, which offer low-cost and accessible legal options for victims who want to get their images removed. That province also has direct support lines for victims and survivors to learn about their rights, but also about the technological and social supports available to them. The BC Society of Transition Houses and Tech Safe BC offer supports for people who want to have their devices examined to see how digital harms against them have occurred. These services are not entirely perfect. They're limited to specific harms, but they're a good start.

Across Canada, these types of services, government-funded programming and fast-track options are lacking. We need well-formatted laws and broad supports, including direct frontline victim services with expertise in technological issues and well-resourced arms of government and civil society organizations that engage in meaningful consultation, research and education to understand these harms as they evolve, assess the effectiveness of the legal remedies that are available and proactively seek legal, social and technological changes to support victims and survivors. This includes holding technology companies to account when they don't create AI programs that are safe or free from discrimination and, rather, act as tools for increased harms.

Thank you for your time.

4:50 p.m.

Conservative

The Chair Conservative Dominique Vien

Thank you very much, Ms. Dunn.

I now give the floor to Ms. Parent for five minutes.

Emmanuelle Parent Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

I am going to take some of my five minutes to start the recording. We all know how social media work.

It is a pleasure to meet you. I have a doctorate in communications and I am the executive director of the Centre pour l’intelligence émotionnelle en ligne, an organization that educates, listens to and equips young people in Quebec at the primary and secondary levels. I will focus on young girls and young people from LGBTQ+ communities in this presentation.

4:50 p.m.

Conservative

The Chair Conservative Dominique Vien

Ms. Parent, I am sorry to interrupt. I am advised that you are not allowed to record while you are testifying.

4:50 p.m.

Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

Emmanuelle Parent

I will stop right there. I respect your right to consent or not.

4:50 p.m.

Conservative

The Chair Conservative Dominique Vien

It is no more serious than that.

Please continue.

4:50 p.m.

Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

Emmanuelle Parent

Great. I am glad I let you know that I was recording.

At CIEL, we have spoken with over 100,000 children and youth since 2020. So today, I am trying to report their opinions and experiences with artificial intelligence.

I was with you eight months ago to talk about anti-feminism. I read your report. Congratulations. Thank you for listening to us. I also read the government's response, which said how it is already responding to the recommendations. I have a more specific message: hear, hear, that is good, but it is not enough. Let us see how we can improve the response.

First, I am going to talk to you about the perception of young people, including girls and 2SLGBTQIA+ youth. During focus groups, when we ask them what they think about artificial intelligence and how they want us to educate them, there is a consensus: They say they are tired of hearing about artificial intelligence. How did we get there? That somewhat illustrates the dichotomy between pro-artificial intelligence narratives and anti-artificial intelligence narratives.

I am quoting a teenager. He told me that he has a teacher who prohibits him from writing with artificial intelligence. Another teacher suggests AI should be used in a healthy way. One teacher corrects his work with artificial intelligence, but another teacher tells him that he is using up too much water when he uses artificial intelligence and that he must limit its use. His mother connected his account to her ChatGPT account, but when he does a search on Snapchat or Google, he uses artificial intelligence. What is he supposed to do exactly?

Between narratives about the development of AI, that it is for everyone and for the future, and fear narratives that AI is going to lead to the extinction of humanity, young people do not know where to go and they lose faith in what we are telling them.

So, our first recommendation is to adopt a balanced approach that addresses both the positives and the negatives when talking to young people, including girls and 2SLGBTQIA+ youth.

First, Canada's national artificial intelligence strategy: AI for all states that, when funding initiatives on artificial intelligence literacy, we must act as a watchdog by ensuring that a balanced approach is always used, that no campaign focuses only on online threats, and that there is a balanced view of both sides.

Then, on top of that, there is the stigma surrounding digital device use. Young people will say that they do not want to seek help, because it is yet another screen problem. They are told that they spend too much time using their phones, even though we adults are the ones who buy them phones. In Quebec, more than 80% of 12-year-olds have a smart phone.

However, when young people fear being judged, they are reluctant to talk to an adult. Who can they turn to? Why would they not turn to AI? According to a U.S. study by Common Sense Media on 1,060 teenagers, over 72% of them have already used an AI companion. About one in three teens even said they prefer to talk to AI rather than to a person face to face. We wonder why.

I will quote another teenager who opened up to us. According to him, artificial intelligence does not judge him. The risk of being judged is something that girls and teenagers from 2ELGBTQIA+ communities are all too familiar with. Tools such as Character.AI and Replika, or simply generative artificial intelligence, do more than just answer questions. They can simulate emotions in relationships and become the people to whom young people turn to discuss their most private concerns. When the questions become increasingly personal—for example, when discussing one's gender or sexual orientation—experts say this is when young people are most at risk.

Here is my second recommendation. If we want young girls and 2SLGBTQIA+ youth to turn to adults for resources, we need to secure online spaces where young people will not be felt guilty. Bill C‑34, safe social media act, sets out parameters to ensure that artificial intelligence is subject to greater restrictions in the case of AI companions and chatbots.

However, unfortunately, these platforms are not reliable. Let me give you an example: OpenAI's ChatGPT claims to have a type of user account with settings for teenagers. Researchers at Common Sense Media tested these settings. The teenager was allowed to discuss suicide, self-harm and eating disorders; no notification was sent to parents. Therefore, Bill C‑34 must also provide for audits to verify the measures that platforms claim to have put in place to comply with our laws.

Lastly, artificial intelligence is much more than generative AI; it's everything that governs what girls and young people in 2SLGBTQIA+ communities are exposed to on social media because of recommendation algorithms and content moderation. Young people—

4:55 p.m.

Conservative

The Chair Conservative Dominique Vien

Thank you, Ms. Parent.

4:55 p.m.

Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

Emmanuelle Parent

There you go. I would be happy to talk more about that.

Thank you for listening.

4:55 p.m.

Conservative

The Chair Conservative Dominique Vien

We can see the enthusiasm; there will be some fruitful discussions with the members.

Dr. Kendall, the floor is yours for five minutes.

Dr. Tamil Kendall Director, Partnership for Women's Health Research Canada

Thank you, Madam Chair, for the invitation to appear.

I represent the Partnership for Women's Health Research Canada, which unites the Women's Health Research Institute in B.C., Women and Children’s Health Research Institute in Alberta, Women's College Hospital in Ontario, the Research Institute of the McGill University Health Centre in Quebec and IWK Health in Nova Scotia. Together we are the largest national network of women's health researchers, with 74 Canada research chairs and more than 1,200 members.

The health of girls, women and members of the 2SLGBTQI community is central to their employment, economic advancement and productivity, the topic of this particular study. In fact, McKinsey estimates that closing the women's health gap would contribute $37 billion annually to Canada's GDP. Ensuring that women and gender-diverse individuals are not left behind in the AI health care revolution is essential for closing that gap.

AI systems analyze massive datasets, interpret complex medical images and synthesize research evidence in seconds. This creates an opportunity to accelerate learning in health care in ways that simply weren't possible before, but AI is only as good as the data it trains on. Currently, AI makes existing biases and discrimination in health care worse. This is true for race, ethnicity, sex, gender and sexual orientation.

Let's take women's heart disease as an example. It's the leading cause of premature death among Canadian women, but important historical research studies included men only. Men are still more likely to be diagnosed and appropriately treated than women. AI exaggerates these biases. It's less likely to diagnose women's heart conditions than men's heart conditions, and it's more likely to attribute women's symptoms to mental health issues.

But if we build it and train it properly, AI can improve accurate diagnosis for women.

An example is hypertrophic cardiomyopathy. It's an inherited genetic condition that makes the heart muscle abnormally thick. It's equally common in men and women, fifty-fifty, but two-thirds of the people who are diagnosed currently are men, because the diagnostic criteria, like so many, are based on the male body. Researchers developed an AI tool to analyze MRI heart scans, what AI can do better and quicker than humans can, but they also built it to adjust the diagnostic criteria for age, sex and body size. The number of women diagnosed increased dramatically, to 44%, better aligned with the actual occurrence in the population.

AI can improve women's health and health care rather than compromise it further, but to do so, it needs to be appropriately trained. There's world-class Canadian women's health research happening. It emphasizes that we need sex-specific normal ranges, diagnosis and treatment options. When we do this in our specialized women's heart health centres, women are receiving women-specific care and experiencing better outcomes. This is the information we need to train AI on.

To change the current situation and ensure that AI is applied ethically and effectively, we recommend that the federal government fund the proposed Canada strong for women's health research and data platform, making the research and data from women's hospitals and research institutes from coast to coast available for AI applications. This requires data from patients who are not hospitalized, as 80% of health care is delivered in outpatient settings, and should be led by women's health experts.

The government should implement existing equity, diversity, inclusion and SGBA commitments that have already been made by the federal government and most provinces and territories, such as implementing standards for disaggregated sex and gender identity; going further by improving standardized data collected on race and ethnicity to understand access to and appropriateness of care; and having Health Canada require that AI providers test their systems for bias against specific populations before an AI tool can be licensed for sale.

The federal government should work with provinces to ban the use of commercial AI models that use health care spending or past clinical utilization as a proxy for disease severity. These metrics under-represent women and 2SLGBTQI people and others who face systemic barriers to accessing care.

Finally, as others have said, mandate postdeployment audits—

5 p.m.

Conservative

The Chair Conservative Dominique Vien

Thank you very much, Ms. Kendall. The time we had for your opening remarks is up, but you will have a chance to follow up during the discussion.

Mrs. DeRidder, the floor is yours for six minutes.

5 p.m.

Conservative

Kelly DeRidder Conservative Kitchener Centre, ON

Thank you, Madam Chair.

I have a couple of questions.

First, Ms. Dunn, you mentioned that there are gaps in the current federal government framework to protect women from the harms of deepfakes and non-consensual intimate images. Can you explain a little further where these gaps are and how we could close them?

5 p.m.

Associate Professor and Director, Law and Technology Institute, Schulich School of Law, Dalhousie University, As an Individual

Suzie Dunn

There are some immediate gaps when we look at quick takedown options. Often, people are looking for quick takedown options. They currently exist only in British Columbia and Quebec. The safe social media act, if it passes, will allow for a quick takedown.

One of the gaps that we need to address is a place that someone can call when they're being harmed to figure out what their options are. When we think about a criminal response, yes, that might recognize the harm they've experienced, but when it comes to the speed of the response, there's a delay on the criminal side of things. Especially, with these tech-facilitated harms, that's very important.

The two other big pieces are around the creation and hosting of the content. When we look at purpose-built platforms, like MrDeepFakes and these types of websites, they have no purpose. Their whole purpose is harm. What are the ways the government can respond? This might occur through the safe social media act, but there is also other implementation that we can think of.

When we think about the creation of these apps, are we enforcing privacy laws adequately? Are we looking to the Google apps or wherever these apps are hosted?

There can be things around advertising rules and looking at social media companies that are still advertising DeepNude apps to teenagers and adults when we know these companies are not following basic privacy laws and they're creating illegal content. There needs to be a government response, not just for the individuals who are causing harm to individuals but also for the platforms that are allowing the creation and proliferation of the content.

5:05 p.m.

Conservative

Kelly DeRidder Conservative Kitchener Centre, ON

Thank you so much.

Emmanuelle, you gave a really excellent example of how AI is being used in everyday life, from water tracking to grading, general info lookup and even using AI as a companion. How do you think we can balance the education on AI, specifically about personal protection, while still acknowledging that AI exists in everyday life?

5:05 p.m.

Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

Emmanuelle Parent

That is a great question. In my opinion, awareness campaigns or tools that are developed for individuals will, for example, document only the dangers or risks of artificial intelligence. So, the focus is on the negative. On the other hand, there are programs that will, for example, talk about future jobs in the artificial intelligence sector and women in the world of technologies. It is as if the negatives of AI and the positives remain in distinct silos.

So, what does that mean in their daily lives of young people? They know that sometimes it is good and sometimes it is not. However, we need to be able to have a balanced approach in each of these fields so that when we talk about entrepreneurship, we also talk about the fact that these are imperfect tools because, as we were saying, there is a lack of data on women. When we talk about threats, we also have to say that it can be used on a daily basis. So, I think we need to merge the two narratives about the benefits and about the downsides.

Young people will continue to use artificial intelligence on a daily basis. Even if we ban conversational agents for young people, we have to expect that young people will circumvent the rules to use these tools. This is what we see in other countries. So we still have to remain balanced and open, and we must not judge them.

I hope that answers your question somewhat.

5:05 p.m.

Conservative

Kelly DeRidder Conservative Kitchener Centre, ON

Thank you very much. That was very in depth.

Dr. Kendall, I have one question for you.

Being an expert in women's and children's health, you mentioned how AI will enhance diagnostic systems to better diagnose women in Canada. Can you give some good examples of how that's already happening?

I might as well ask my follow-up question for you now. Do you believe there needs to be a national data framework for our health data in Canada?

5:05 p.m.

Director, Partnership for Women's Health Research Canada

Dr. Tamil Kendall

I think it's really critical. This also refers to the comments made by others. We need to constrain artificial intelligence and direct it toward good sources of data. When we do that, we actually can make changes.

I gave one example of diagnosis around a genetic heart condition. There are other examples of starting to use AI as a predictive model. You could build those models so that they close some of those biases in identifying and diagnosing women.

Another place where AI is going to be incredible is in the cancer space, where you can analyze images so much more accurately. This can really be a game-changer in cervical cancers, gynecological cancers and breast cancers, as well as other types of cancer.

However, we have these data gaps. We don't—

5:05 p.m.

Conservative

The Chair Conservative Dominique Vien

Thank you, Dr. Kendall.

5:05 p.m.

Director, Partnership for Women's Health Research Canada

Dr. Tamil Kendall

I get excited and go on. I apologize.

5:05 p.m.

Conservative

The Chair Conservative Dominique Vien

You will have a chance to go back to this.

Ms. Ménard, you have the floor for six minutes.

Marie-Gabrielle Ménard Liberal Hochelaga—Rosemont-Est, QC

Thank you, Madam Chair.

We are just at the beginning of this study and, as colleagues have pointed out recently, we can say that we have entered a third revolution. Society went through the Industrial Revolution and then, an automation revolution happened. We are now in an artificial intelligence revolution. I think that we, the members of this group, have a moral duty, but we also have to engage all parliamentarians.

I go back to the title of the study: The impact on women and the LGBTQ community is clearly being felt.

I would like to hear your comments, witnesses. An expert pointed out to us that there are perhaps two key observations to be made at this stage. Firstly, artificial intelligence is riddled with bias. It is based on pre-existing content. Secondly, not only are women and the LGBTQ+ community under-represented or portrayed through the lens of bias, but there is also an under-representation of these individuals among those who develop artificial intelligence. Is there anything the government can do to help women and people from the LGBTQ community get involved in the development of artificial intelligence in order to counteract what we are seeing?

What are you seeing in terms of the development of the various platforms?

5:10 p.m.

Associate Professor and Director, Law and Technology Institute, Schulich School of Law, Dalhousie University, As an Individual

Suzie Dunn

The big question of data is hard, because we have history behind us. I think some of the data questions were around funding more research for women's and LGBTQ issues more generally.

When it comes to having this technology dealt with and guided by people who come from feminist and queer perspectives, there's a funding gap. When we look at the stats on venture capitalism, very few women's and LGBTQ products are funded. Even if women and LGBTQ people have great ideas, they can't get the money for them. If you're thinking about the ways we can have people engaged in these processes, there is a funding gap. Can the government encourage scholarships? Can it encourage tax incentives and these types of things for these types of products?

The thing is that we have to be careful about this. It is not just that women and LGBTQ people will make good products. You need people who are not just representative of their identity, but who are also creating products that are good for women and good for LGBTQ people. This is where, if we're thinking about having things like the digital safety commissioner and other government agencies that are going to be assessing the adequacy and the safety of products, you need to have good representation of people who care about issues like equality, autonomy and dignity in assessing those products.

5:10 p.m.

Cofounder and Executive Director, Le Centre pour l’intelligence émotionnelle en ligne

Emmanuelle Parent

I would like to take this opportunity to mention that young people could be consulted on the initiatives and projects that are to be funded. It is interesting to get their feedback and consult them not only before we decide which projects to fund and which bills to introduce, but also afterwards, regarding the impact of these projects and bills. Whether the issue is digital literacy or artificial intelligence, we should keep young people at the centre of the debate.

We hear about the whole issue of representativeness. But when it comes to artificial intelligence, I think what is most represented are the economic interests of the platforms seeking to make a profit. These interests should be balanced with the collective interests that you all cherish here, in a very important way. So I think young people, too, could have a seat at the table.