Introduction

This report presents the findings of the Survey of Online Harms in Canada 2025, the sixth national study conducted by the Dais to examine Canadians’ experiences with social media platforms, exposure to online harms, and attitudes toward regulation and solutions.

The report begins by situating the findings in the current regulatory and geopolitical context, outlining stalled federal legislation, evolving global frameworks, and recent rollbacks in platform safety practices.

Next, the report explores the online media environment in Canada—tracking how Canadians use social media platforms, where they get their news, and how those trends vary by age and political orientation. This section also assesses levels of trust in news sources, tech platforms, and digital services.

The heart of the report analyzes Canadians’ exposure to online harms, including misinformation and disinformation, synthetic media, hate speech, fraud, and harassment. It reveals which groups are most affected and highlights the growing impact of generational divides in media consumption and harm exposure.

The final sections examine what Canadians think should be done about these harms. This includes their views on individual responsibility, and the perceived (in)effectiveness of tools like blocking and reporting. Our findings reveal Canadians’ strong support for system-level solutions. Potential solutions range from fact-checking tools and platform accountability measures to content labelling, bot removal, and age-based restrictions.

Methodology 

This study was conducted from January 20, 2025 to January 27, 2025 with Pollara Strategic Insights. In total, 2,502 Canadian residents aged 16 and older were interviewed, using respondents provided by the online panel provider Léger. The sample of 2,502 Canadian residents was weighted by age, gender, language, and region based on the latest Census data, to ensure the sample is representative of Canada as a whole. Results are compared against previous iterations of the study conducted by the Dais since 2018. The complete survey instrument and demographic breakdown of the sample are available in the Appendix, along with methodology information for previous iterations of the study. Totals presented throughout may not sum or add to 100 percent due to rounding.

Online Harms

Canadian residents are exposed to a wide range of potential harms through their use of digital platforms—from exposure to fake news, to hate speech, identity fraud and beyond.

Figure 13 shows that the most common harm that respondents notice in posts, links, images, or videos on online platforms is false news (i.e. misinformation). Three-quarters of Canadian residents have spotted news or current events that they immediately suspected as false at least a few times a year. Nearly as many reported seeing false news they initially believed to be true and later discovered to be false (71 percent). Similarly, 67 percent have seen synthetic media or deepfake images or videos online. Exposure to hate speech against an identifiable group and identity fraud are both nearly as high, while exposure to speech promoting physical violence is slightly lower with just over half of Canadian residents (51 percent) reporting having seen it at least a few times per year.

This section will dive into Canadian respondents’ exposure to each of these harms online and highlight the groups most vulnerable to or at risk of each harm.

Disinformation and misinformation

The top category of online harm experienced by Canadian residents is false information about news and current events. Generally, “misinformation” is any false information that spreads, without any intent behind it. This is contrasted with “disinformation,” which is any false information intentionally spread to cause damage to individuals, organizations, countries, or any other group. While misinformation is usually unintentional, it can still harm because it spreads virally on social platforms. Figure 13 shows that 12 percent say they see false news multiple times a day and a large majority are exposed to it more generally.

This has consistently been true over time (see Figure 14). While it was largely steady between 2022 and 2024, there has been a slight increase in its prevalence. While in 2022 only 32 percent saw false news (and instantly identified it as fake) multiple times a week, that has since risen to 38 percent in 2025. Similarly, in 2022 only 16 percent of respondents believed fake news at least a few times a week but that has grown to 23 percent now.

Respondents who said they had seen examples of false news online were asked to describe an example of what they had seen. The most commonly cited examples relate to US politicians, and in particular US President Donald Trump—a topic that experienced a significant jump from 11 percent in 2024 to 22 percent in 2025. Conversely, examples related to celebrities fell from 22 percent in 2024 to 16 percent in 2025.

Because this survey was conducted around the time of the Los Angeles wildfires, this event featured prominently since false information related to them was circulating online. This included fake photos of the Hollywood sign burning, or misattribution of blame, and other stories that were false—with some cases like the fake burning Hollywood sign likely being disinformation (circulated with the intent to deceive), while other stories simply being misinformation spread more naturally. This is an example that reflects the speed at which misinformation and disinformation circulate for breaking news events. There is nothing about the wildfire disaster that made it particularly susceptible to false claims—all current events face the same risks of false information spreading quickly.

Table 1

Most common topics of false news seen online

As noted, the more pernicious form of fake news online is untrue information that Canadian residents actually believe—this comes in the form of misinformation or disinformation. While this distinction is important, for the purpose of this research we only assess belief in false information, regardless of whether it was spread intentionally or not.

To measure belief in false information (or misinformation) we ask respondents to assess whether or not eight different statements are true (see Figure 15). These statements reflect a range of beliefs that have circulated online recently—from climate denial to 15-minute cities (the belief that government is intentionally limit the free movement of Canadians), to COVID-19-related conspiracies. The list also included one statement that has been flipped so that the statement is true and saying it is false is a sign of belief in misinformation as a control (the statement “the number of natural disasters is increasing due to climate change”).

Aside from the flipped statement, belief is highest in a version of the “great replacement theory” (the false belief that there is an intentional plan in place to replace native-born populations in Western countries with immigrants). This is a more recent development; when the statement was first tested in 2022, only 16 percent said it was either definitely or somewhat true (see Figure 16).

That has now risen to over one-quarter of Canadians (28 percent) including 10 percent who say it is definitely true. This rise has coincided with significant public discourse in Canada surrounding immigration over the past year.7 The next most common belief is that climate change is a natural phenomenon. While it has grown since 2022 from 22 percent to 25 percent now, it has grown more slowly than the great replacement theory.

While belief in the other statements has remained consistent over time, there has been some softening in beliefs surrounding the war in Ukraine. In 2022, 73 percent of Canadian residents identified the statement “Ukraine nationalism is a neo-Nazi movement, so Russia invaded Ukraine to protect people” as untrue. That figure has dropped to only 66 percent in 2025 with the share of residents unsure rising from 20 to 25 percent.

We group Canadian residents together based on the number of statements they correctly identify as not true (or true in the case of the single flipped statement). This means that respondents who are unsure about any statement were treated as having not correctly identified that statement. The result is a scale ranging from 0 (no statements correctly identified) to 8 (every statement correctly identified).

Doing this, as shown in Figure 17, we find that just over half (52 percent) of Canadian residents fall into the “low misinformation” group which includes anyone who scored at least 6 out of 8. A further 35 percent fall into the middle group, scoring between 3 and 5, and the remaining 13 percent of Canadian residents are classified as the “high misinformation” group—correctly identifying at most two statements out of the eight total.

This represents a slight but statistically insignificant increase in overall belief in misinformation since 2024. We cannot compare this against previous iterations of this survey due to changes in the statements used.

Belief in misinformation is not spread evenly across the population. In general, older, wealthier, and better-educated Canadian residents tend to score higher on the misinformation index while younger, lower-income, and less well-educated Canadian residents do worse (see Figure 18). This has been consistently true over the years of this survey.

While there are not massive regional differences, Quebec has consistently stood out as a region where residents tend to perform better on the misinformation index. In Quebec, 58 percent scored at least a 6 on this scale, 6 points higher than the national average.

Consistent with both past results from this survey and work by other groups like EKOS8, those on the right of the political spectrum tend to believe more misinformation than those on the left.

We also can compare this misinformation index with where Canadian residents say they get their news. For this, we use the average number of correct statements (which for the entire survey population is 5.23).

In general, respondents who say they get news from traditional sources such as news websites, news alerts on phones, the radio, and other similar traditional media sources tend to have the highest average number of correct responses. Those who read news websites in particular have the highest overall scores on the misinformation index. See Figure 19.

While most social platforms tend to be associated with worse scores, Reddit and Threads both have users who score above average on the misinformation index. While Threads is new to the survey this year, Reddit has consistently been an outlier from other social platforms.

Other social platforms do worse on this measure. WhatsApp has the lowest average score, but all the most popular platforms including Facebook, Instagram, YouTube, and TikTok score below average on the misinformation index. There is a general trend that relying on social media networks as a source of news is associated with higher rates of belief in misinformation.

ChatGPT also falls at the bottom of the list. This is consistent with last year’s results, despite the fact that ChatGPT as a platform has added new features to its base model allowing the tool to pull in up-to-date information. However, this data does not allow us to disentangle whether the platform causes the low misinformation index score or if those who are already prone to misinformation are more likely to use AI tools like ChatGPT.

Figure 19

Average number of correct misinformation statements by news source

Deepfakes and synthetic media

Recently, concerns about the increasing ease of creating deepfakes have become front and centre in the minds of many Canadians, as well as policymakers concerned with the integrity of informational ecosystems. With cheap or free generative AI tools capable of generating increasingly realistic images, video, and audio easily available and becoming better with every passing year, Canadians are now seeing more synthetic media than ever.

Now, 67 percent of Canadian residents report seeing deepfakes or synthetic media online at least a few times a year—up from 60 percent of Canadians just last year (see Figure 20). This includes 30 percent who now report seeing synthetic media multiple times a week (and nine percent who are seeing it daily).

While most of the conversation surrounding deepfakes focuses on their threats as it relates to political misinformation, the most common type of deepfake that Canadian residents report seeing is related to celebrities. A close second, however, are deepfakes of Canadian politicians.

Table 2

Most common subject of deepfakes seen online

Hate speech, fraud, and harassment

Canadians’ exposure to hate speech, identity fraud and impersonation, and promotion of violence have all become more common online over the past three years. Although the overall number of Canadians reporting exposure to these harms has remained largely stable (outside impersonation), the frequency of exposure has increased.

Figure 21 shows the share of Canadians who say they see hate speech against an identifiable group at least a few times a week has grown from 22 percent in 2022 to 27 percent in 2025. Similarly, the share who see content promoting physical violence has also grown from 15 percent to 18 percent in that period.

When it comes to identity fraud or impersonation, there has been growth both in overall exposure and in frequency of exposure. In 2025, 22 percent reported seeing this type of content multiple times a week compared to 19 percent in 2022, but overall exposure to fraud at least a few times a year has also grown from 59 percent to 63 percent.

While most Canadians are exposed to these harms online, fewer report personal experiences of being targeted. In 2025, 12 percent of Canadian residents say they have been targeted by hate speech online, eight percent have been targeted with online harassment causing them to fear for their safety, and five percent report having had intimate images shared online without consent (see Figure 22).

The incidence of targeted hate speech is significantly higher for many equity-deserving communities. For racialized Canadian residents, recent immigrants, those living with disabilities, and 2SLGBTQ+ residents, hate speech is anywhere from 50 percent to 100 percent more common (see Figure 23). Women experience slightly lower rates of targeted hate speech.

The topic of online hate speech has changed significantly since 2024. Last year, online hate speech related to the Israel-Palestine conflict was most prevalent. In 2025, the most cited topic of online hate speech is exposure related to newcomers to Canada, with many specifically citing hate speech toward Indian immigrants. Israel/Palestine remains a significant topic of online hate speech reported by Canadians (17 percent) with many identifying exposure to the related topics of antisemitism (eight percent) and Islamophobia (three percent).

Table 3

Most common topics of hate speech seen online

Solutions

Given the prevalence of harmful online content on social media platforms, Canadians are seeking solutions. In this section, we explore who respondents believe is at fault for the rise in harmful content, how interventions to address harmful content are working, and what types of system-level policy interventions they believe governments should pursue.

Attribution of fault

Canadians consistently report that users of social platforms are most at fault for causing the rise in harmful content. In 2025, 43 percent of Canadians say users of social platforms are most responsible for causing the rise in harmful content, while only 20 percent attribute fault to the platforms themselves and 11 percent to government or political leaders. A quarter of Canadians do not think anyone is responsible or are unsure of who is to blame (see Figure 25).

When asked who is responsible for fixing it, these numbers are reversed. Nearly half (47 percent) say that online platforms are responsible and another 21 percent say that government or political leaders are responsible. Only 14 percent say that users are responsible for fixing the rise in harmful content.

Individual responsibility

Users are taking a number of actions to address their own exposure to harmful online content and help improve the state of social media platforms.

Blocking and reporting

The primary tools available to users online are blocking and reporting accounts and content they deem harmful.

Four in ten Canadian residents say they have blocked, reported, or flagged an account for being either fake or automated online. Another quarter have either reported or flagged, or blocked an account for sharing illegal content (see Figure 26). A smaller number of Canadian residents (eight percent) say they are also reporting illegal online activity to police.

However, Canadians believe the impact of these tactics is limited. Only 37 percent of those who have blocked, reported, or flagged a user online rate this action as effective (and only 10 percent as very effective). Comparatively, 14 percent say it was not at all effective and 29 percent total called it ineffective. See Figure 27.

Fact checking

When it comes to misinformation, disinformation, and deepfakes, users are often told to fact check what they see online to make sure it’s correct. As noted above, 70 percent of Canadian residents have seen false information online but immediately recognized that it was false. However, fewer Canadians report that they fact check what they see. Figure 28 reveals that only 61 percent of Canadians say they have ever fact-checked something they see online, despite the persistent warnings not to believe what you see online. More concerning, those who believe the most misinformation are also less likely to fact-check online content. Among the group that is more likely to report belief in misinformation, only 40 percent report ever fact checking information using other sources—more than 20 points lower than the 61 percent of Canadian residents who report fact checking.

System-level

Given the perceived limitations of user-side solutions to harmful accounts and content and the responsibility that Canadians attribute to platforms and governments to address these issues, we also asked about attitudes toward regulatory interventions for online platforms.

Platform regulation

In general, respondents support government action to regulate online platforms to limit online harms, even when presented with tradeoffs. As Figure 29 shows, two-thirds of Canadians (68 percent) believe that reducing the amount of hate speech, harassment, and false information online is more important than protecting freedom of expression. Similarly, two-thirds believe the government should intervene to require online platforms to act responsibly and reduce the amount of harmful content on their platforms and believe government should intervene to reduce the intentional spread of false information because it is a threat to Canadian democracy. These attitudes have held steady since 2022, with support for all three pro-regulation statements consistently between 64 percent and 70 percent across all three waves of the survey.

Figure 29

Attitudes toward online platform regulation in Canada

Canadians' attitudes are not consistently aligned with either pro- or anti-regulation positions. Only about half of Canadians agreed with all of the pro-regulation statements, while only about 15 percent agree with every anti-regulation statement. The remaining third of Canadian residents agree with some pro-regulation and some anti-regulation statements.

When asked about specific interventions that place requirements on online platforms to address online harms, Canadians are overwhelmingly supportive.

Figure 30 presents a range of potential regulatory requirements for online platforms, with nearly every proposed approach supported by more than three-quarters of Canadian residents. The highest levels of support are for requirements that aim to safeguard children or limit the spread of non-consensual intimate images. The proposed intervention with the lowest level of support (still 65 percent) would allow the government to order platforms to take action during times of crisis, including blocking or promoting specific content. Levels of support have been consistent or growing over time (see Figure 31).

Banning TikTok

Canadians were also asked about whether they support banning TikTok in Canada, in response to concerns about the potential risk of surveillance or interference by the Chinese state given the company’s ownership structure. The Government of Canada has already banned the use of TikTok on government-issued devices, and the United States has taken significant steps toward banning or forcing the sale of the platform (although, more recently, that has been pulled back).

As shown in Figure 32, a majority of respondents (52 percent) support banning TikTok in Canada, with only 19 percent opposed and 29 percent who are either neutral on the issue or don’t know yet. This is largely consistent with our findings in 2024, when 57 percent of Canadians supported either fully banning TikTok or banning it for users under 18 years of age.

Unsurprisingly, given the demographics of those who use TikTok, younger Canadians are significantly more likely to oppose a proposed ban than older groups. Among respondents in the 16 to 29 age group, only 32 percent support banning TikTok while 35 percent are opposed to it. At the opposite end of the spectrum, only nine percent of respondents over 60 years of age oppose banning TikTok while 67 percent support it.

Conclusion

The findings of this survey continue to present a paradox. Canadians’ use of social media platforms like YouTube and Facebook is nearly ubiquitous and has been rising on an array of newer platforms. Canadian residents’ media diet increasingly consists of news and information from social media rather than directly from news organizations themselves—with some of the most used platforms, run by Meta, actually blocking Canadian news content.

Yet Canadians’ levels of trust in online platforms—notably the Meta platforms, X, and TikTok—is extremely low. The people who use these platforms for news and current events are more likely to be exposed to false and misleading information than those who use traditional media sources. And Canadians continue to report high levels of exposure to harms like hate speech, identity fraud, and non-consensual sharing of intimate images. Exposure to these harms is more common for younger Canadians, as well as for other marginalized communities. These harms are likely to be amplified by synthetic media and deepfakes, produced by new and rapidly advancing generative AI technologies.

What is to be done to address these online harms on social platforms?

Canadians believe that platforms should be primarily responsible for fixing these problems. Some Canadians are employing user-side tactics offered by platforms, like blocking and reporting harmful content and accounts to the platforms, or independently fact checking the sources of information they are seeing. Yet a plurality of respondents do not perceive these tools and tactics to be effective. And those with the highest levels of belief in misinformation are least likely to use them.

Canadians strongly support the regulation of online platforms. Even when offered a tradeoff, about two-thirds of respondents want governments to regulate platforms to reduce the amount of hate speech, harassment and false information; to require platforms to act responsibly to reduce harmful content; and to address the intentional false spread of information that threatens democracy. This has been consistently true for the years we have tracked these attitudes.

When presented with specific policy interventions to impose requirements on social platforms, Canadians are even more supportive. The highest levels of support are for requiring the removal of the most egregious types of harmful content, such as child sexual abuse materials and non-consensual sharing of intimate images. But there is also strong support for increasing parental controls, quickly removing fraudulent accounts and bots, addressing hate speech, labelling synthetic media and deepfakes, and many others. No proposal has less than 65 percent support from Canadian residents.

In sum, Canadians are very clearly calling for government action to address online harms.

For appendix, see PDF page 57