Defending democracy: institutional responses to online misinformation

Earlier this year , Canada's public inquiry studying foreign election interference released its final report and it held a stark warning .

I'm your host , takara Small , and this podcast is brought to you by CIRA , the Canadian Internet Registration Authority , the nonprofit building a trusted internet for Canadians . We've got some great guests in this episode , but first up is Sue Gardner . You may remember Sue from our first episode . She's a journalist and tech commentator who used to run CBC Online .

Among many other things , sue has been telling us all about the scale of the task facing the government in regulating this space .

So Bill C-27 , you know , was aimed at , in part , ensuring that AI technology got used responsibly . So there are implications there . Or potentially , c-63 , probably the big one , right ? Online harms would have imposed a duty of care on the large social media platforms .

That would have had implications for hate speech , for extremism , for terrorism broader implications , probably , too .

It also proposed to establish the Digital Safety Commission , which would have been a sort of a center for expertise , thinking about this stuff and grappling with it and trying to get out in front of it , not just misinfo and disinfo , but including misinfo and disinfo . All of that got washed away , right when the government was prorogued , so all of that is gone .

It would need to be brought back again if wanted , right ? And so the answer is we're not really anywhere , really , and the only thing that's comforting about that , I suppose , is that that is true for most everybody , right ? I mean , that's not entirely true . Right , the EU , germany .

So Canada , as a result , has kind of just , you know , on the sidelines taking a look at what's happening in the EU , taking a look at what's happening in the US and kind of hedged their bets , right , and I feel like that is very much writ large .

And as the US veers off into whatever it is , whatever road it is going down , we are going to have to , as a country , obviously like figure out what do we do now , right , if we can't just be on their coattails and sort of a mini me version of them . What therefore , are we and how do we pick it and how do we make our way to it ? Right ?

So it's a bigger question too .

There's been , like there's been , a petition I don't know if you saw it , you know asking Canada to consider joining the EU and to me what that symbolizes the fact that once again Canada doesn't necessarily , or Canadians don't necessarily , feel like Canada can be on its own .

It has to have a leader of sorts it needs to have someone directing where to go , how to feel about certain issues . And you know , a petition , even in jest I mean it had hundreds of signatures , but even in jest says a lot about , perhaps , the role that Canadians feel they should play on the global stage .

It's not of a leader , it's maybe of a quiet supporter .

If we don't have that , we're very , very small , right . So I think a realistic appreciation of our heft is probably not a bad thing .

So how do you go about regulating misinformation in a an online world where being able to to convey your feelings and opinions is taken as a freedom ?

Who decides what is okay versus what is outside the Overton window and not okay ? How do you avoid government overreach ? How do you avoid ? Oh , I like it today with this government , but tomorrow , with that government , it's going to be very oppressive and terrible .

Right , like those are real , obvious , time-tested questions , right , but I think that what we're forgetting when we have those conversations and some of us are forgetting this on purpose I think that what we're forgetting is that that all needs to be understood in context .

And in democratic countries today , in liberal democracies , today , we have a hilariously larger amount of free speech than we have ever had before . Right , like everybody got handed a printing press , everybody got handed a microphone . Right , it is today .

Now it is totally okay to shout fire in a crowded theater , and we have people doing it every single day , all day long , right , so it's a different landscape . You know , in a lot of ways these changes have been super fantastic , right , because you know it used to be

that there were a lot of people shut out of the social discourse for a variety of reasons , right , some people were like invisible to the gatekeepers . The gatekeepers just didn't know about them or care about them , and some people the gatekeepers disagreed with . Those people now get to speak , and that is great and that is real and it is new .

But it's also true that we have never had 100% , entirely unconstrained freedom of speech , even in the United States , right , it's just not a thing Like . There are libel laws , there are truth in advertising laws . Lots of countries have hate speech laws .

You know , we have always been historically very careful during election campaigns because there's such a time of vulnerability , etc . Etc . Etc .

And so here we find ourselves , right where most of the world you know we talked a little while ago about the EU , right , like most of the world has taken a more moderate approach when it comes to freedom of speech than the United States .

And here we find ourselves with the limits of our speech being largely set by American social media companies , unless we take steps to make it otherwise . Right , they have been setting the terms of the discourse and they have also been setting the terms of the discourse about the discourse , right .

So Silicon Valley has kind of set the stage , they've set the table , they've set the terms of the discussion right .

And so now , suddenly , newly , you know , all of the world is operating in a much more US-like environment , but you know , hyper-US environment when it comes to freedom of speech relative to where we were previously , and again , with the caveat that some countries have been advancing Germany especially right , advancing to impose their own cultural norms on internet

platforms , right . And so that's the context that I feel like that needs to be understood in . We have so much more freedom to speak and to be heard , right , and , generally speaking , that is good .

But also , and equally importantly , we have always constrained our speech and you know the normal trajectory , have always constrained our speech , and you know the normal trajectory . Technology makes it possible . We run off and do things , harms emerge and then we constrain those harms . That's normal .

And so I feel , like you know , we're having these conversations about misinformation , freedom of speech .

In some ways we're having them as though it's the first time we've ever had them and like they're shocking and horrible to us , like we're furious about them , and I don't understand that , because I feel like these are extremely normal conversations that we have always had right . We've always had them . Like Canada , right .

We made hate speech illegal a number of years ago and then we decided to not have it be illegal and then we decided that we're now going to consider again having it be illegal , like . Those are good conversations . That's a healthy society talking to itself about what it wants to be like right and what the trade-offs are of various approaches .

I think that's good .

Commissioner Marie-José Ogue was in charge of it and when she announced her findings she had this to say Foreign actors are no longer content to use traditional means to interfere .

harmful . The impact of traditional methods should not be underestimated , but the greatest threat the one that I believe threatens the very existence of our democracy is disinformation . This threat is all the more nefarious because the means available to counter it are limited and very difficult to implement .

Nevertheless , we must not give up , but rather attack it forcefully altogether . This cannot be the sole purview of government . It will not work .

Helena has a particular focus on large language modeling and how that's being used to bring disinformation to a wider scale .

Also , on a level of processes , canada is targeted , for example , like election interference , but also in this context it's very important to know that on an individual level , there is also constant targeting of decision makers .

Even researchers like me who study disinformation , and the journalists who are doing the job of debunking disinformation and actually enlighten the public on that issue are targeted . Activists are under constant attack for this job that they are doing .

Maybe you're interested to know , and I think it is important to set the stage and explain why authoritarian countries and anti-democratic regimes are doing it . There are several kind of strategic goals that foreign countries seek to achieve through such interference , active measures or information warfare , whatever you call it .

It is the purpose and the strategic goal is the same . This includes , first of all , polarizing public debate to inflame these divisions in the society . Such topics could be like vaccination , immigration , indigenous issues and how they work . They just can landscape for this existing division and looking for these weak spots and just amplify them .

There was a really cool research a few years ago that showed that Russia state-affiliated media significantly amplified coverage of the Freedom Convoy protests in 2020s , publishing information about this event 15 or even 20 times more often than outlets like BBC , cnn or Deutsche Welle .

And this amplification actually aimed to artificially increase the significance of this event and therefore deepen the polarization among Canadians . Another goal of the authoritarians is to erode public confidence in democracy and institution by convincing that the public institutions , these very public institutions , are no longer fit for purpose , and that's how they saw .

In this internal distrust , there are a lot of also external goals , for example , to weaken Canada's alignment with allies in NATO , in G7 , but like that's a whole other different topic . And what also important to understand with that regard is why Canada is actually a target and also will be continuing to be a target for us Rotariansussians .

Because , first of all , due to the strong democratic tradition of our country , which Austro-Russians regimes will never be okay with and will find it actually threatening . Similarly with it for example , invasion of Russia in Ukraine they were not okay having the democratic neighbors .

That's why it will always be a threat and , ironically , these non-democratic regimes exploit Canada's democratic nature it's freedom of speech , like academic freedom and other fundamental rights and using the very freedom against our country .

Research also , like our colleagues did , incredible research that shows that they are weaponizing radical left and radical right in Canada to undermine support for Ukraine and seed other narratives .

And unfortunately , which is like really pity , but democratic nations are limited in how they can respond to that threats because asymmetric or like covert retaliation would actually compromise the very nature of democracy which is Canada .

Second reason why Canada will be a target it is that actually Canada is considered one of the most covertly infiltrated liberal democracies in the world .

Fragmentation of information ecosystem where all these national communities are hanging out , engaging the type of content they are consuming , and this actually provides authoritarian actors with very , very fertile ground to meddle in Canada's internal affairs , in Canada's internal affairs , weaponizing information channels and messengers , social media , to orchestrate this influence

campaign .

And the last point that I want to mention with that regard , why Canada will be a target , is that Canada very closely integrated with the US information ecosystem because of the same language , because of the same cultural space , same stars , bloggers , influencers , and it also makes Canada vulnerable to the interference that is happening into the US by other countries .

And I don't know if you remember this huge uh scandal that happened last year that it involved um in the us , uh two rt employees and they were indicted for covertly funding and directing uh company tenant media , actually funded by um Canadians for publishing thousands of videos for Kremlin , like in the interest of Kremlin , and research by Media Ecosystem Observatory

it's a laboratory at McGill . They were actually analyzing the content of these bloggers and how it intertwined with Canadian informational media ecosystem and the findings were absolutely incredible . First of all , one quarter of Canadians were aware of Tenet Media , affiliated creators or bloggers .

They were engaging with their contents and the other findings showed that among 2,000 podcast episodes that they were relentlessly analyzing , almost a quarter of them like 500 , mentioned Canada and very often in highly polarized and negative tone , and topics for discussion were frequently criticizing immigration policies , social justice movement and like reinforcing these divisions .

So it's also kind of like that spillover effect on canada . So it's like a long answer for a short question , but it is very important to fundamentally understand why Canada is targeted and will be targeted and why authoritarian countries are generally doing this .

First of all , a side task force which stands for security and intelligence threats

to elections . It is a task force which is like a whole government working group responsible for assessing and sharing intelligence during the election period . It also consists of experts from CSIS , communication security establishment , global Affairs Canada and RSMP . And RSMP there is also critical election incident public protocol .

It is a panel of senior public servants empowered to alert the public about these incidents that actually could threaten integrity of federal elections . So they're doing kind of this pre-banking work which is really important . I'm sure your listeners know about rapid response mechanisms under Global Affairs Canada .

It's like mechanisms , strengths and G7 coordination in identifying and responding to foreign threats to democracies . They are sharing information , doing joint analysis and also trying to find some opportunities for coordinated responses together with other democracies . There is also really called Digital Citizen Initiative .

It's like a multi-component let's say strategy , and they are supporting digital literacy . Counter-information counter-component , let's say strategy , and they are supporting digital literacy , counter-information , counter-disinformation projects .

Among NGOs , I would personally love to highlight academic initiatives like Canadian Digital Media Research Network and I already mentioned that from the perspective of media ecosystem observatory , because media ecosystem observatory is running this network .

They are supported by canada heritage and do an absolutely incredible job in fortify for a fortifying resilience within a Canadian information ecosystem , trying to identify some vulnerabilities and doing amazing analysis that helps policymakers to make a decision and actually help other as just ordinary citizens , to understand what is happening in the informational ecosystem .

There's also communication security establishment and CCS . These are agencies that are actually responsible for threat assessment , cyber monitoring , and they do provide Canadian government with IT security and foreign signal intelligence and that sort of stuff .

Additionally , there are several legislative frameworks , as you're probably aware of and we'll probably talk about that a little bit later , like Bill C-26 , bill C-27 , c-63 , all of them are working to somehow address the issue of informational integrity and security on the internet .

In my latest report for CG , which is about AI and information manipulation , I'm analyzing Russia's interference in the US election very , very narrow and particular case . I analyze how AI was used by Russia to conduct disinformation campaigns in the US and I used a dataset from Atlantic Council DFR lab and categorized the incident using FIMI topology .

So this research gave me a really cool understanding on how AI has been used for information manipulation . But after that , some really cool development in that field happened , but I will talk about that in a moment . So what did I found out in my reports that the most common tactic was using a 65 percentage was a narrative manipulation .

So AI was used to craft and distort narrative by producing fabricated news stories , images , videos , sometimes even testimonies of fake whistleblowers , to , to , to create this false and misleading content . Create this false and misleading content the most powerful .

Actually , these examples were combining real and synthetic material and it's like a very old-school Soviet-style approach where they were combining real material and forged to deceive people . So not something super new for us , but using these new technologies technologies , these tactics are evolving . So in some cases , actually , what research pointed out ?

That audio fakes were more compelling than video ? In others , even simple , cheap fakes were even more efficient than AI fakes . The second types of tactics were identity falsification . It was in 55% of the cases . This involved using AI to forge identities , create fake personas or like mimic trusted sources to deceive audience .

Maybe you've heard of this um operation called doppelganger in this operation . russian actors used ai to deliberately create this believable clones and copies of reputable news outlets in the EU and then in the US and recently Canada also was targeted by this doppelganger operation , so it is related to identity falsification .

Other group , like , as I identify them , was , like ai for amplification , so use artificial intelligence to automate content distribution and manipulate social media algorithms . This tactic relies on ai and actually ai and bolts working together in tandem to boost the reach and visibility of this manipulative content .

And the last category is strategic targeting , where influence operations were precisely tailored to specific audiences , like with the languages , with some even local variation of different languages , and for specific demographics . So like apparently , ai enables these campaigns to be very adaptive , responsive , efficient and , what is really concerning , super cheap .

But after all this like analysis finished super recently , like one month ago . We all learned about , learned about llm grooming . Have you heard of that ? I haven't . So that's a really cool phenomenon , uh , so , uh , russians deliberately feeding content into large language models , so they were creating the network called pravda .

It's series of websites that are not actually designed for human consumptions . They're not really user-friendly , they don't have this search bar and articles don't really resemble typical journalistic content . It's just pieces of some telegram posts and other content and they actually post like 2 , like 2000 of them in a day .

Generally , these sites cumulatively produced 3.7 million articles in the past year . So why do they do that ? Their primary goal is to seed training datasets of big , large language model , because the chatbots are scraping the internet looking for new content to kind of educate LLMs , and that's how finally they are getting into your output in ChatGPT .

So you cannot be sure that next time when you're trying to get some information from chat GPT it will not be Russian propaganda . So they're kind of legitimizing their narratives through this very , I need to admit , creative way to do disinformation .

It's actually the end of the democracy and that's what happened in Russia . Russians do not believe in any kind of truth because they do understand that everything could be disinformation . So democracy doesn't die with a coup . It dies very slowly by erosion and from the inside , as citizens lose a face and their participation in democracy actually declines .

So it's become the very easy target for non-democratic regimes and then you can actually find some parallels with real-time events right now with your neighbors , because that's actually what is happening and it is existential . So in its core , actually , democracy really require like a shared set of facts so that people can debate solution and not debate realities .

So when misinformation spread unchecked , it create multiple and like incompatible realities and people are starting living in that realities and it's become dangerous because what is true become contested and it is actually eroding democracy itself .

But if we're talking about general , comprehensive strategy definitely like a long-term strategy , like building resilience , inoculating people against disinformation definitely more could be done and it cannot be a case for complacency not yet swayed . It doesn't mean future immunity , because actors , as you just heard , are very creative in their approaches .

They will be using new and new technologies , new tactics and new vulnerabilities will be emerging in this fight . So it's very good finding that elections have not been swayed yet , but more could be done .

insights . Now Helen Hayes is a senior fellow at the Center for Media Technology and Democracy at McGill University , and she joined us to talk about the government's approach as well . Helen is an expert on government policy in the media ecosystem . So how has the federal government been responding ? With legislation .

That was meant to regulate social media platforms and the types of content available on social media platforms , but we currently don't have any past federal legislation that actually does that work . This is in part because C63 was tabled and then died on the order paper after Parliament was prorogued , and so now we're kind of starting from scratch here .

We don't have a regulatory framework for mis and disinformation specifically , but even for social media and the internet generally .

So we're kind of starting from ground zero with this new government that was just elected and we'll have to see what they do to try to address it , based on , in part , the Hogue Commission's most recent summary report about the importance of addressing mis and disinformation in order to safeguard Canadian democracy .

We also saw , in the case of C63 particularly , that a lot of the discourse related to its tabling became highly partisan . So we saw conservatives saying that digital policy was an attempt to regulate people's speech and infringement of people's free speech .

We saw lobbying powers from big tech saying that this type of regulation would disincentivize investment from big tech in Canada , from big tech in Canada . And then we saw members of the Canadian civil society coming out against various aspects of C63 because they felt like they didn't have a good place in that piece of legislation .

So , for example , we saw with C63 , not only a set of core provisions like the duty for platforms to act responsibly , the duty to protect children and the duty to make certain content inaccessible , but we also saw an attempt to amend parts of the criminal code and parts of the Human Rights Act to address hate speech online , and those two parts of the bill

became highly politicized , highly contentious and in my view and I think in a lot of other folks' views that stifled the ability of C-63 and its core provisions that duty to act responsibly , that duty to protect children , that duty to make certain content inaccessible really difficult to pass , even though in my view doing so would have definitely safeguarded Canadians

and Canadian children online more effectively , because right now obviously there isn't any regulatory mechanism to do so there isn't any regulatory mechanism to do so , but Canadian institutions are also limited in their power , right , because so much of the problems we're seeing are spread on platforms that are based outside of the country .

And how can we effectively regulate these companies ? Well , what's interesting about the approach that Canada has taken previously is its alignment with other places around the world . So we think about alignment with the UK , a big powerhouse , a first mover on online safety . Alignment with Australia and its e-safety commissioner .

Alignment with the EU broadly , and this makes up kind of this global nexus of power that incentivizes big tech to pay attention to regulation and to run their companies based on a more aligned regulatory model that makes sense for their business .

It also means , from a Canadian perspective , that the government can say we need to protect the Canadian internet infrastructure , we need to protect Canadians and what they see online , and to do so we're going to incentivize big tech .

Companies have to be held accountable for ensuring that young people aren't subject to self-harm content that's algorithmically fed to them on their feeds .

We need to make sure that we're setting up those guardrails , and I think we could do that predominantly with regulatory alignment in the global context , so thinking internationally rather than siloing Canada as its own market , looking to regulate as part of this broader regulatory alignment structure .

That makes sense and makes us more powerful because we're working in numbers against big tech .

. I often look at the uk as a first mover and a leader in this space , particularly when it comes to protecting young people online . So you may have heard of a thing called the age appropriate design code , that that was originated in the uk , and it calls for a systems design approach , which I think is a really effective approach to digital regulation .

It basically calls on platforms to ensure that their systems are designed in a way that upholds rights and digital safety for the youngest users on their platforms . So this doesn't mean stifling people's speech . It doesn't mean censoring content . It means setting up the system itself to work for the benefit of young users . I think that's really effective .

This can happen through things like turning off location services for young people , about allowing young people to have more control over the content that they see , allowing young people not to be in contact with older users on on these platforms . So I think the UK , in terms of child safety , would definitely be the the leader of the pack .

Australia is right up there , though , when we think about the eSafety Commissioner and the way that they are viewing child safety online , which I think is a really key component to this . This . They actually just announced a device ban , which obviously is contentious .

It does , in some respects , align with what we're seeing in Canada with device bans in schools , but we see kind of these regulatory mechanisms that are aimed at supporting children's well-being , and I think that when we're talking about , you know , broader concerns about the information ecosystem , we really need to be talking about how kids are affected by the spread

of misinformation , for example , or by the content that they're algorithmically fed , because they're our next generation , they're the ones making moves in the political arena , they're the ones who can influence policy , and so I think those issues are really inextricably linked and we need to pay attention to that when we talk about digital regulation generally .

I mean , in the UK , a lot of the advocacy related to this type of legislation came from parents who saw their young children being drastically affected by their use of online platforms and , in particular , there were a series of incredibly tragic stories with young people taking their lives as a result of the content that was fed to them on their social media feeds

, and this led to a groundswell in advocacy and activism related to digital governance . So how can we actually hold big tech to account ? Because we know that big tech can have the ability to distort discourse or to change our views or to shape our society . How do we hold them to account in a way that protects the young people that we're losing ?

And their activism and advocacy was incredibly strong on this file , both in austral and in the UK , but I would venture to say the UK was the most strong , and that was led by a series of civil society organizations , academic researchers and , like I said , parents organizations , children's rights groups who saw these issues the issue of internet use and social media

accountability with the health and well-being of their children , and that made a huge difference to the passing of the legislation .

We see that with the Media Ecosystem Observatory , for example . We see that with DisinfoWatch . We're seeing more of a critical approach to discourse and narratives that circulate online . That I think is really important for a healthy democracy .

That , of course , means that academic researchers need access to data , which is a huge hurdle in Canada , and we , I think , need to work on those transparency provisions in order to allow researchers to do that important work . I think they're critical . And then obviously , advocacy organizations and we see that with parents groups in Canada particularly .

We need to have kind of a whole of Canada approach to this where we view digital literacy , digital policy , tech policy as an important integral part of all the work that we do , because really it affects the state of our democracy , it shapes our society , and so I think we can look at all sectors and all places , not just with young people but also community

organizations that can help create a more digitally literate and engaged citizenry , so that they know , when they're encountering information online , how to engage with it , how to understand whether it's factual or not , and what organizations to trust , what to be more skeptical of and how to form your own opinion based on the information that you take in from online

spaces .

So the DCI , it's an initiative of the government that is basically founded founded on these ideas that Canadians need to and should have access to diverse but also reliable sources of news , and that's because , obviously , canadians should be able to form their own opinions , they should hold governments accountable and individuals accountable , and they need to know how to

do so through reliable sources of information and news that allow them to create their own opinions .

So the DCI basically supports these ideas about citizen resilience against disinfo or misinfo by empowering organizations , civil society organizations and academic organizations to either do research or produce programming that can support Canadians in achieving a healthy relationship with their online information ecosystem and also , in the long run , kind of protect democracy against

ills related to mis and disinfo , foreign interference and other dubious content online .