00:00:21.000Learn how at Ontario.ca slash Safer Ontario. Paid for by the Government of Ontario.
00:00:30.000If there's one thing that the Kearney government is focused on above and beyond, well, expanding trade internationally and building out Canada's economy, it seems to be regulating the Internet.
00:00:44.040Hello, I'm Brian Lilly and welcome to the Full Comment Podcast.
00:00:46.960This week, we take a look at Bill C-34.
00:00:49.880Its full name is An Act to Enact, the Digital Safety Act and the Digital Safety Commission of Canada Act.
00:00:56.500Its, well, short-form name would be the Social Media Act.
00:01:00.920But it's a lot more than just online harms and social media for kids, putting in a ban for those under 16.
00:01:07.880What is this act all about, and are there concerns about it?
00:01:12.500Joining us to talk about this is Emily Laidlaw.
00:01:15.160She is the Canada Research Chair and Cybersecurity Law Professor at the University of Calgary.
00:03:43.160this particular one has been consulted extensively.
00:03:47.400I co-chaired the expert group to the government. There were citizens' assemblies, youth assemblies,
00:03:53.900consultations across Canada that culminated in Bill C-63. But as you mentioned, there was a lot
00:04:00.420of controversy about that bill. 90% of the bill was about this online harms of regulating social
00:04:07.980media platforms. Most of that was not controversial. So the big pushback was the fact that they decided
00:04:15.180to make an omnibus bill and tack on amendments to the criminal code, reintroducing the Human
00:04:21.360Rights Act. And that is where most of the controversy was. And of course, Parliament
00:04:26.740was prorogued in January 2025, and that bill died. So this is kind of the second bill, Bill C-34.
00:04:34.620It is the third effort, and it is unlike the predecessor, it just reintroduces platform
00:04:43.420regulation. So there's no attempt as previously you mentioned the Canadian Human Rights Act
00:04:50.320and a long time ago many of us fought a provision that was in there that allowed you know we
00:04:59.080basically said it was the hurt feelings regulation that someone could complain to the Canadian Human
00:05:04.620Rights Commission and get a judgment a financial judgment against someone because they felt that
00:05:11.420they'd said something that was harmful to their feelings. Not wrong. Mark Stein famously went0.97
00:05:19.000through all of that. There was a big push that was gotten rid of. There was an attempt to bring
00:05:22.780that back in the previous version, wasn't there? There was an attempt to bring it back and to keep
00:05:27.320it more focused, though. So some of the provisions in it were really tailored to trying to limit
00:05:33.800the kind of frivolous and vexatious complaints that you're speaking to.
00:05:42.180But there wasn't popular support for it.
00:05:45.760And I find that I'm in rooms where it's really mixed.
00:05:48.820And I also teach an anti-hate course at the University of Ottawa
00:05:53.400where we study this and the students are really mixed.
00:05:55.800We dig into freedom of expression law.
00:05:58.600And so that one is just, I see no indications of that being reintroduced by the government.
00:06:03.800I still laugh at one of the reasons that Mark Stein was, for his piece in McLean's, was cited as pushing hate was he made a joke about Islam.
00:06:15.280He said, is it just me or does Ramadan seem to come around earlier and earlier every year?
00:06:20.140Well, that only is funny if you know anything about Islam.1.00
00:06:23.920And Ramadan comes around about every 11 and a half months.
00:06:26.800So, yes, it comes around earlier every year because it's not a 12-month holiday.
00:06:33.460And that was used back 20-odd years ago to say, well, you're promoting hatred.
00:06:38.620I'm glad that they're not going that way.
00:06:40.760But do you feel that in the current situation, have they struck the right note or are they going too far?
00:06:49.200And I'll use one of your expressions that I love.
00:06:53.180Are they still protecting lawful but awful expression?
00:06:58.420Yeah, and that is the big question is the lawful but awful, right?
00:07:01.300So I will say this. And if you'll indulge me, maybe I'll explain how the law works a little bit. The new law strikes the right balance. Let me put aside the social media ban, because I think we have to have a separate conversation about that.
00:07:17.300The gold standard now is this idea that these enormous, incredibly influential businesses, these big tech companies, should have consumer protection responsibilities and that they should be mitigating the risks of harm, right?
00:07:35.480That essentially there is a business model here that can cause harm and they should be regulated.
00:07:41.140At the moment, there is very little regulation of what they do.
00:07:44.760So in a way, this is what kinds of harms, though.
00:07:48.940So there's two types of harms that I think are a real focus.
00:07:55.380So these social media is essentially designed to be like the sticky design features, right?
00:08:01.600If you're looking at the algorithms that are pushing certain repetitive content, we're seeing with kids pushing of self-harm content, pushing of eating disorder content and sticky design features.
00:09:09.800And so they feed you material that will keep you on their platform. That's what you mean when you're talking about the sticky content. Now, I want people to go to my material on my website. They want to keep you there. And so they're going to feed you things that will give you that dopamine hit.
00:09:25.980It's the dopamine hit, but it's also likes and angry emotions and putting in little rewards that if you keep using something, you might achieve a certain level and get some sort of gift, right?
00:09:44.440And there's a lot of gaming involved, particularly for kids.
00:09:49.600And so it's how they design these spaces.
00:09:52.600It's basically this is the business that they're in.
00:09:54.720This is how they keep people using their products so that they make money. That's fine. It's just it's a consumer protection question about is leaving it to the companies to make decisions about how to craft that protecting consumers or does that need there need to be legal intervention to to put in place certain base business standards.
00:10:15.540And that is what this type of law, at least on a whole, targets.
00:10:20.440The other thing is that there's some extraordinary harm that we've seen that's flown from them.
00:10:25.760And this act targets social media and chatbots.
00:10:28.500So we can talk about chatbots in a bit.
00:10:32.120And let me talk about kids, although it's not just about kids.
00:10:36.580What we're seeing is that increasingly as kids are using these spaces,
00:10:41.260We're seeing a huge increase in exposure to essentially strangers that are exploiting these spaces for criminal ends, whether it's grooming and luring kids and then threatening them and sextorting them.
00:10:57.880And oftentimes they're connecting with them in real life.
00:11:00.480We have some horrific instances where, you know, a kid was lured through Spotify and then kidnapped.
00:11:06.380We are seeing extremism play out in these spaces.
00:11:09.340For example, they'll often befriend these children and then they'll get them to do certain things and then they'll threaten them with it. And they become basically drawn into an organized crime racket that is for extremism purposes.
00:11:24.500And this is real, and this is happening on the front end. And we are seeing a huge proliferation of child sexual abuse images, intimate image abuse of children and adults. Well, the adults for the intimate images. And now the creation of AI-generated synthetic images. So there's a lot going on there, right?
00:11:44.840At the moment, most of the laws focus rightfully on the wrongdoer. Of course, you can't find them half the time. And you're also looking at quick ways that the spaces and the platforms that are these gatekeepers in the middle should maybe have business obligations to try to handle these particular harms.
00:12:03.800And that's what the legislation is aimed at. Of course, the lawful but awful that you mentioned, that is that key in-between space, which is when are platforms overreaching and being forced to overregulate sort of the day-to-day conversations that might be offensive, that might be hateful, but don't rise to the level of being something that would be criminal or otherwise unlawful.
00:12:31.740I will say that the legislation intentionally does not, for the most part, target that, and we can dig into some of that if you want.
00:12:41.560So, look, we've heard about incidents of young people, kids, harming themselves for many years.
00:12:49.920I think back to the Amanda Todd case, and that has to be 15 years ago, maybe?
00:13:44.740So is what you're talking about, the negative harms, the children being groomed, whether sexually or for organized crime,
00:13:53.000Is that a common thing? I'm not saying like one out of three kids or anything like that, but is it a very small group or is it bigger than most of us can imagine?
00:14:06.180It is. I mean, it's that's a hard question to answer. I would say like all crimes, you know, most people don't murder, but it's still a crime. Right.
00:14:15.640So it is not happening to most kids, but it is happening to more than we think. And I think that there's a tendency to view it as, and I have these conversations all the time, which is, oh, but my kids are fine. They're okay. They're not into anything problematic.
00:14:32.740And I think all kids are at risk. Some are more at risk. I think marginalized youth in particular are more at risk to this kind of extortion. But the number of stories about kids who are not marginalized that have faced this situation, I think the question we have to ask is a bit different.
00:14:53.520And it is, is this a whole of society problem? And should there be certain safety measures? And safety measures don't mean always content removal, right? So it might be, hey, your kid is on this space and maybe if you figure out who the kids are, you shouldn't allow them to live stream or have private chats because we know that some of those private chats are where this grooming takes place.
00:15:20.860Or it might be things like we should have more curated algorithms so it's safer for children.
00:15:27.560What does the legislation do to deal with this, though?
00:15:31.300I mean, we'll talk about the social media ban in a moment.
00:15:34.640But beyond that, you know, kids are going to end up or they'll be 15 one day and 16 the next and then they can go on.
00:15:42.360And so what does the legislation do to change algorithms or to alter what people can see without blocking me from being able to see the content that I want that is lawful and hopefully not awful?
00:16:00.440And so this type of legislation, I say it's this type because we're seeing something similar in, for example, Europe, which is it takes a scalpel approach and it creates it's you create a regulator that's acting in the public interest.
00:16:17.000And the regulator in the case of Canada would only have the power to order content removal when it's child sexual abuse and intimate images.
00:16:24.540So stuff that we'd all agree should be taken down.
00:16:27.740And for the rest of it, for things like fomenting hatred, inciting terrorism and all the child protection areas, companies need to put in place risk mitigation measures and they need to be transparent about it.
00:16:42.280It basically is you need to lift the lid on the business model of these companies.
00:16:46.520And it's not about individual pieces of content.
00:16:49.220It is about what safety measures do you put in place?
00:16:52.420What are your content moderation practices?
00:20:03.720That is a false sense that you ban it.
00:20:06.080That means that once they're onto these sites, they're automatically safe and that's not necessarily the case or that they work. And so your kids are kept safe. But we know kids are devious. They get around it. I got my hands on all, you know, alcohol before I was vagued. Like we many did.
00:20:22.880I know no that never happens yes this is our confession podcast and so I so my worry with
00:20:32.480the ban is that it's going to create a false sense of safety what it could do though is precisely
00:20:38.340what your friend mentioned which is it becomes it prompts a culture shift so not necessarily
00:20:44.240the kids that already have accounts now are going to try to get around it but like the seven-year-olds
00:20:48.340and the eight-year-olds where it's just not accepted culturally to use social media until
00:20:53.600you're 16. Let me add a wrinkle to this. Canada's proposal is a middle path. It doesn't actually
00:20:59.840propose a ban. What they're proposing is what I would call a pause, which is it's very in-between
00:21:07.740an approach. It says if you're a large social media provider, you're one of the ones we regulate,
00:21:13.460you don't get to access the market that is children
00:21:16.900until you prove to us that you have safety standards in place.
00:21:21.700And if you want to apply for an exemption, you can.
00:21:59.680How much of this is parental responsibility?
00:22:03.880Yeah, parental responsibility is always going to be a key part of this.
00:22:07.040I mean, we see that with the ban, because if they allow their kids to use their accounts or help them, you know, circumvent it, that's an issue.
00:22:14.740At this stage, though, social media and just big tech in general, this is so far beyond parents.
00:22:24.040This is a massive corporate industry that hasn't really had that much regulation.
00:22:32.120And so this is a way to address that and fill that gap.
00:22:37.800There's always an important role to play for parents.
00:22:40.680But I do receive that response sometimes saying, oh, parents should just do a better job at parenting and be involved.
00:22:47.840And it's like there's nothing that they could do, though, to keep their kids safe and make these better spaces for their children beyond a few key steps.
00:22:59.620And that's not even going to go far enough.
00:23:01.260and well we need to take a break but when we come back i will ask you you know do we as adults even
00:23:09.180know the platforms that the kids are on because you know we're talking about facebook well kids
00:23:15.440aren't on facebook we're talking about instagram facebook is for the olds um instagram is for the
00:23:21.080middles you know i mean in probably a lot of them are looking at tiktok and saying oh that's for old
00:23:27.400people. Now, there are so many different platforms that I think people just don't realize. So we'll
00:23:32.040talk about that in a moment when we come back. We've got merch before we break. I want to tell
00:23:38.220you about this. We've got merch from hoodies to hats to tote bags and more. And you can find it
00:23:42.940all. Shop for your official gear, whether it's National Post, Toronto Sun, your favorite post
00:23:48.440media brand. It's a bold way to show your love for the fearless journalism. And it looks good,
00:23:53.540two. I'm just looking right now at the Toronto Sun ball cap. It says Toronto Sun established 1971,
00:23:59.620a great year because that's when I was established. Visit postmedia.store to shop the collection
00:24:06.060today. You're a political junkie, right? I mean, that's why you're listening to this podcast.
00:24:11.960But if you're not in the loop with Post Media's VIP political hack, then you're not getting even
00:26:37.120Do they even fully understand all the different platforms?
00:26:40.440because it seems like, not to sound like an old school guy talking about kids in their rock and
00:26:45.920roll orchestras, but it's like we can't even keep up with all the different platforms.
00:26:53.160Well, no, we can't. And I work in this area and I'm constantly learning about new platforms that
00:26:58.640I didn't know about. And I think that that's one of the challenges. And so let me frame it
00:27:03.740a different way, which is how do you write laws about this then? Our expert group constantly
00:27:10.300had conversations about how do you keep this practical? How do you build out a regulator
00:27:15.540that's just appropriate, but doesn't take on the world? How do you write a law that's flexible for
00:27:21.960the future and the evolving threat landscape? I mean, when we first all met in 2022, we were
00:27:27.300having zero conversations about chatbots. And now that's a major source of tech facilitated harm,
00:27:34.800Right. So one of the things that you have to do is figure out who you're you're aiming at.
00:27:41.920The legislation targets large platforms.
00:27:46.180So a lot of these smaller, newer ones that are coming up that kids might get on would not be regulated by this, except if the government designated them or the regulator, at least, as of a significant risk of harm.
00:28:01.720The two that I just mentioned, Roblox and Discord.
00:28:04.300I mean, my kids who are now grown and young adults, I mean, they're on Discord playing games.
00:28:10.680You hear about Discord showing up occasionally in criminal investigations.
00:28:14.000I would hope something like that might be included in the legislation.
00:28:42.560And so there's a few ways to think about this.
00:28:44.860If you're trying to be practical and you are standing up a new regulator, you've got to keep it narrow.
00:28:50.780And so in that spirit, they really kept focused on large platforms and ones where you can communicate to the world at large.
00:28:58.440Like you get on Instagram, you get on X, and it's capable of communicating to the world at large.
00:29:03.960If it's not capable of that, all Discord servers have a max, they have a cap.
00:29:11.140Roblox, you're not talking to the world at large, it's a game, and games are excluded in this.
00:29:17.140Private messaging like Snapchat, the private messages back and forth, all of that is currently excluded.
00:29:23.880The way that they dealt with that was to create a category called online services.
00:29:28.480And basically, in the future, you could add different types of technology as the regulator's ready, as you see that there's a significant risk in those areas.
00:29:39.920And that they, you know, when it comes to kids, the governor and council could add it.
00:29:43.940So it wouldn't be going through parliament and debated.
00:29:46.480I give you all this detail because choices have to be made.
00:29:49.740I look at it and I think some were very disappointed and I wanted to see gaming added to that space. Criticisms are, though, you don't want to include all gaming. Not all games are a threat. We're really talking about things like Roblox. I would like to have seen something like Discord included. But how do you figure out the caps on something like that?
00:30:10.060So these are choices and they're hard ones.
00:30:13.840And I think that people should have strong views and express them about what the government should be doing with these decisions that they made.
00:30:22.120The ability to communicate in games is far and above what it was for my kids when they were young and playing on Club Penguin, which, by the way, great Canadian company that got sold and then got shut down.
00:30:32.500But it was a fun game while it lasted.
00:30:35.400And so I'm surprised that they didn't go in that way.
00:30:38.000But there are critics, such as the Justice Center for Constitutional Freedoms, I want to read off to you part of what they preambled with their petition asking the government to stop Bill C-34, which hasn't passed.
00:30:53.820There's a whole bunch of bills that did pass at the last minute.
00:31:50.980I think that there are some important questions that they're asking, though.
00:31:54.660The setup of the regulator itself and the appointment of the commissioner is pretty typical for a regulator. It's the same as the Competition Bureau, for example. So the CRTC is the same. So it is independent. If we want to push for something even more independent from government, yeah, absolutely. That should be something that's advocated for.
00:32:18.600I don't see how it undermines parents rather than reinforcing it.
00:32:23.980And I know that that's an argument that has been made that we should just leave this to parents and it's undermining parental choice.
00:32:31.020One avenue for something like the ban is to say parents should be able to provide permission for the kids to be able to use these spaces.
00:32:41.620I will say most feedback from parents had been, dear God, no, don't like ban it, but then put it on me that I'm still the one having to navigate.
00:33:22.560There's always the devil in the details about, you know, how you look at the gray areas of offensive expression that isn't actually illegal.
00:33:32.320They have raised the point about privacy, though.
00:33:34.460And let's talk about that, because for social media bans to work, they need to use age verification. And that's the privacy argument, that it forces social media platforms to put in place some mechanism to figure out who the kids are so that they can then ban them, or including age verify them to ensure they're not accessing pornographic material.
00:33:59.320So there are really important privacy and cybersecurity questions about age assurance. The bill itself doesn't specify the method of age assurance, but it does require that the companies put in place best practices that protect privacy and cybersecurity.
00:34:19.160I think many of the arguments out there see all age verification as just a non-starter. It's an automatic privacy invasion and it's no-go zone. I don't agree. I think that there's a lot of developments with age assurance. I think there's a risk. So the risk is that a company puts in place sloppy solutions, collects our IDs, knows who everyone is. It makes it vulnerable to hacking. We've already seen that hacking happen.
00:34:44.900but in practice there's a lot more trade-offs here there are a lot of choices how would you do
00:34:51.640age verification though um yeah so i mean so that if you're whether it's saying kids under 16 can't
00:34:59.100go uh on instagram or we need age verification for a pornography site i mean do you have to
00:35:07.020scan your driver's license if you're under 16 you don't have one do you have to scan your id
00:35:58.100That's something I hadn't thought of until now.
00:36:01.480Well, and that's a really good question.
00:36:03.640I actually don't know the answer to that one because I haven't thought through the connections between Bill C22 and these companies.
00:36:11.980I would say that for the most part, that wouldn't be required because the act itself actually requires data minimization and best practices developed through the privacy commissioner.
00:36:21.360I'd also say around the world, they're pushing for age-appropriate design for kids.
00:36:27.460Well, to do that, you need to know who kids are.
00:36:29.560So the money question in the next five years is figuring out who kids are.
00:36:33.240But let me give you examples of other ways that age assurance works.
00:36:37.140I've had a Facebook account since 2007, so I'm over 16.
00:36:42.520They might look at your Gmail account and be like, oh, you've had that for 20-odd years.
00:36:49.280Now, the less reliable measures are kind of age-estimating, where they look at the faces of the kids and the behaviors of the kids, and they're like, I guess that you're 12.
00:37:02.080because you look 12, you've just announced that you graduated elementary school and you're like
00:37:07.460TikTok video, fairly certain you're not supposed to be on here. That's less reliable. And so
00:37:13.200Australia's implemented a waterfall approach to say there's different tiers and you have to think
00:37:19.260of them as kind of different reliability and different privacy risks. So it is not automatic
00:38:43.260Bill C-34 introduces obligations which are so alarmingly broad that providers of regulated services will be tempted to overcomply at the expense of users' freedom of expression and privacy rights.
00:38:58.600I can see that happening where companies just say, well, I don't know.
00:39:06.200So let's just err on the side of caution and we're going to take this approach.
00:39:11.320Yeah, it's always been a concern in this space. And I'll say that law I mentioned from Europe back in 2000, their initial approach, and it is still law in Europe, is that these platforms have to, if they're notified that they're hosting any illegal content, they have to take it down or they risk liability.
00:39:30.400And so that incentivizes companies to err on the side of caution and just start taking down anything, right? That looks vaguely illegal. That isn't the approach taken in this legislation, but I take heart what is being said about the free expression point.
00:39:47.520So the law says that it leaves a lot to the regulator, and that's been criticized, right? I think it's the only way this is really going to work, but the commission, the Digital Safety Commission, will develop regulations and guidelines.
00:40:03.880They are required when they develop those to do so, taking into account freedom of expression and privacy. So when they detail how companies go about compliance, they set out that they should be ways to balance out to ensure freedom of expression is met.
00:40:21.740One thing that can be strengthened in this law, and I have pushed for this for a while, was that the company should have obligations to part of their digital safety plans should be reporting on how they're respecting fundamental rights. This is part of the European law. I think it could be part of the Canadian law.
00:40:41.300So the regulator will be required to enforce it on the companies. The companies up front aren't necessarily required to take that into account. Throughout the legislation, there are comments saying you are not in any way required to disproportionately impact freedom of expression and how you comply with this.
00:41:01.600that that might be viewed as too woolly. I think that it likely is the only approach that then
00:41:08.840kicks it to a regulator to detail. But you also have to have faith in the regulator to do so.
00:41:16.240And I think that will depend on who is appointed. I was going to say my my mileage when it comes to
00:41:22.180regulators does vary. And if you get someone who is overzealous, well, then.
00:41:28.480you know, it could head in the wrong direction. This is just such an area where you've got to
00:41:41.820tiptoe really. And you mentioned you lived in Britain for a decade and they had laws going
00:41:48.760back to 2000. You know, we've all seen the stories of people being arrested in Britain
00:42:21.300Do we have anything in this bill or other bills that the government has put forward recently that approach that zone?
00:42:32.480No, the law in Canada bears very little resemblance to the UK besides the kind of core idea that companies should be responsible.
00:42:42.980So the UK law is much broader, and it takes on all kinds of illegal content, including some, I think, of the more controversial clauses about commenting on, I don't remember the exact phrasing, but there is a clause about immigration, for example, which has become a hot-button issue.
00:43:04.640And UK law, even before the Online Safety Act, they had all kinds of laws that were for kind of the olden days where poison pen letters were written that were repurposed for the digital age that were used to essentially show up on someone's doorstep because of something that they posted on Facebook or X back in the day.
00:43:27.860And there are many instances of that. So that tradition in the UK about what I would say is at times over regulation of expression, that isn't in this law. So the law is focused on seven categories of regulated content and is for the most part, they reflect content that is criminal or criminal adjacent.
00:43:50.680I think where it departs is more for child protection or in the area of chatbots, kind of concerns about the companionship side of chatbots.
00:44:01.180So none of those other features you saw in the UK would happen here.
00:44:05.360And so if we take hate speech, which is one of, I think, the more challenging areas when you're talking about illegal versus lawful but awful content, there would be no obligation for this regulator or no power for the regulator to order content takedown.0.56
00:44:21.760What it allows is for the regulator to investigate a company if it becomes obvious that at a systemic level, they have not been handling content that fements hatred. So they can investigate companies for that. And I think they should. And I think some would be disappointed that there's no avenue to take down what might at times be very obvious criminal hate speech. But that is the way that the act is structured.
00:44:50.240You mentioned chatbots, and there was a lot of discussion after the horrific incident in Tumblr Ridge, trying to lay blame on a specific chatbot, saying that, well, this is why the shooting happened.
00:45:04.680again i have to say is this a moral panic because in that instance we know that the perpetrator in
00:45:14.240that case had been in contact with police numerous times had been sent to a mental health institution
00:45:20.840social services were involved and in the end the government appeared to be blaming
00:45:27.720a chat bot in saying, well, they asked about this. Well, the person had clearly expressed
00:45:34.360similar sentiments to live humans and we didn't act. And so are we looking to make
00:45:43.480technology a scapegoat here? I'm not letting the AI company in question off the hook,
00:45:53.720But there were an awful lot of people before the chatbot was engaged that could have stepped in and said, oh, hold on a minute.
00:46:00.700We've got something bad happening here.
00:46:13.580We know that the perpetrator, she was posting concerning comments on on ChatGPT.
00:46:25.440And so if we think anything is private about using ChatGPT, it's not because those concerns then were flagged up to their trust and safety team.
00:46:33.660And based on their policy, they did not report it to law enforcement.
00:46:37.200And so that is a question of should these companies report to law enforcement when they see concerns? I mean, they're allowed to. A lot of companies haven't placed policies about that. Was the threshold wrong?
00:46:52.260Like, well, clearly we know now in retrospect, this was a real danger. And if they could go back, they would have reported to law enforcement. But I think you can have obligations on the company and also have a look at law enforcement's role that there were failures by different entities when how they handled this.
00:47:13.880Let me give you a different example, though. So it's not just Tumblr Ridge. If we look at companion chatbots like character.ai, there was –
00:47:24.540These are people looking for friends or romantic relationships or what have you.
00:47:29.680Yeah, and I will say I've tried out character.ai on myself, right? I said I was 20 and I wanted to have a friend and wanted to flirt. And like suddenly you're getting pushed all these different sort of romantic exchanges.
00:47:44.920But for a 14-year-old boy, which was this case, it was a pretty horrific result. And the problem with the companionship, it is becomes don't talk to your parents about that. I'm, you know, come meet me on the other side, essentially encouragement.
00:47:59.400And some of the problems with general chatbots like ChatGPT have been how-to guides of how to take your life and, again, encouraging individuals to not seek help from family.
00:48:11.680And so there is this interaction with the chatbot that is not a risk for most people, but is a risk in some instances.
00:48:20.640And so this is that question of safety standards.
00:48:23.660A lot of these companies are developing, you know, putting a lot of work into some of these safeguards, but not all of them. So the question you have to ask is, are we OK just leaving it to the companies to decide for themselves how to address this?
00:48:39.660And I land on the side of no, because I have spent decades watching as they have implemented inconsistent safety measures and sometimes really lax commitment to trying to address this.
00:48:53.040And sometimes they just get slapped around no matter what they do. Right. So I have some sympathy, too.
00:48:59.020So I think that there needs to be some safety standards for and safety safeguards for AI chatbots.
00:49:04.980What I'm really glad about is that they didn't ban them because there was discussion about that.
00:49:11.060And I think you could perhaps consider companion bots, although I'm not for any ban for chatbots, but you absolutely could never have banned just access to general purpose AI like ChatGPT.
00:49:24.140I'm glad that they're not trying to ban chatbots at all.
00:49:27.360I don't use ChatGPT a bunch, but I use several others on a daily basis.
00:49:31.960It's, you know, if you use it right, it can be a great tool for doing what you do in a day.
00:50:16.900There are so many different bills that this government has put forward or championed that touch on the Internet.
00:50:24.260Are we in danger of over-regulating the Internet?
00:50:29.120When you look at them cumulatively, are we headed in that direction?
00:50:34.800I would say back in, when was this, 2018 or 19, the liberal government had published their digital charter, which wasn't really a charter.
00:50:46.840It was just these are the things that we want to address.
00:50:48.900And they didn't address all of those commitments, but it was a suite of reforms all about digital policy. I would say we 100% need to look at our laws to make changes to address the current issues that we're seeing.
00:51:11.020I think it is a fair question to ask whether each individual law is achieving the objectives or is going to have unintended consequences.
00:51:24.120And those are really important debates.
00:51:25.600I mean, some easy ones to my mind would be things like, I think it was Bill C-16, and I think that just received royal assent a few days ago, which adds deepfakes to the criminal coprovision for intimate image abuse.
00:51:44.700Far more controversial would be Bill C-22 and the questions about how long data should be retained, what kind of access should be provided for law enforcement.
00:51:57.880Is this really requiring backdoor access or undermining encryption or is it not?
00:52:03.740These are all really important debates, and I have been in rooms that seem to argue all kinds of different sides of that particular one. So I don't think that they're doing too much, but I think that each one needs to be interrogated closely, and then the determination is whether that individual piece of proposed legislation is just going too far or unnecessary.
00:52:27.340I mean, you never want to introduce a law that you don't need. Bill C-34, we need this one. It is, I teach my students basically a course about non-law because we have very little in this space. But other areas, I mean, you could ask yourself, okay, do we need some of these amendments when it comes to hate speech? Is the criminal code good as is? Is there something missing? Does this have unintended consequences?
00:52:51.280Emily Laidlaw, Canada Research Chair in Cybersecurity Law at the University of Calgary.