The purpose of this report is to analyse the role of digital platforms in contemporary media environments, including public perception of the benefits and problems they bring, especially when it comes to news and information about politics.

Overall, we find evidence for what we call ‘platform ambivalence’. This refers to the fact that many people use platforms for news and information about politics, while also remaining sceptical of the information they see there, and concerned about misinformation, bias, privacy, freedom of expression, and tech power. At the same time, people also appreciate the wider societal and personal benefits, such as easy access to information and staying connected with friends and family.

The report focuses on five platform types: (i) social media, e.g. Facebook, X (formerly Twitter), Instagram, and TikTok; (ii) search engines, e.g. Google, Bing, and Yahoo!; (iii) video networks, e.g. YouTube, Vimeo, and Dailymotion; (iv) messaging apps, e.g. WhatsApp, WeChat, and Facebook Messenger; and (v) generative AI chatbots, e.g. ChatGPT, Google Gemini, and Perplexity.

Based on survey data from eight countries, we find the following, grouped into four categories.

Findings relating to the use of platforms for news and information about politics:

• Across eight countries, search engines are the most widely used platform for news (45%), followed by social media (41%) and video networks (30%). Messaging apps (15%) and generative AI (7%) are not widely used for this purpose. Online news websites and apps (59%) and TV (57%) are still more widely used than any individual platform type.
• The use of news outlets and the use of digital platforms to get news and information are highly intertwined, not an either/or. Most respondents – between 52% and 82% depending on the country – report getting their political news and information from platforms and online/offline news outlets, in part because platforms host content from news outlets. Only a small minority of 10% or less rely exclusively on platforms or, more commonly, only on news outlets.
• People who use platforms for news and information about politics are more likely to also use online news from broadcast and print brands. For example, among those who do not use social media for news and information about politics, 22% use a newspaper website – this rises to 39% among those that do use social media.
• There are important differences between those who use platforms for news and information about politics and those who do not. Younger people, men, those with higher interest in politics, and those more comfortable with technology are all more likely to use platforms for this purpose. Differences by left/right political ideology are small.

Findings relating to the relationship between platforms, politics, and democracy:

• Most people are sceptical of digital platforms as a source of news and information about politics. Just over one third trust video networks (37%), and just under one third trust messaging apps (31%). Social media is trusted by 30% and generative AI by 27%. Search engines are an outlier, trusted by a slight majority (55%).
• Platforms tend to be more trusted as a source of political news and information by younger people and, to a lesser extent, men and those on the political right. These patterns can vary by country. For example, search engines are more trusted by those on the left in Germany, Brazil, and the USA.
• Asked about a range of functionalities related to the public sphere, averaging across eight countries, clear majorities think that platforms have made it easier to connect with friends and family (66%), find the information they need (63%), and connect with likeminded people (59%).
• Clearly illustrating the ‘platform ambivalence’ described above, there is also widespread recognition of possible problems – more than half think that platforms have made it easier for people to say things they wouldn’t say in person (69%), spread misinformation (69%), harass or threaten others (66%), or spread extreme viewpoints (64%).
• Less than half typically think that platforms are systematically biased towards certain political views – but many people are uncertain. Perceptions of bias can be much stronger among certain political groups, with those on the right in the USA (69%) and on the left in Argentina (71%) much more likely to think that social media is systematically biased.
• Asked about what, on balance, brings people together rather than drives them apart, we see large differences across platforms, and between platforms and other actors in society. Messaging apps (+19) and search engines (+12) are platforms that, people think, on balance, bring us together. In contrast, on balance, respondents think that social media divides us (-13). Comparing platforms with other actors in society, many more people think that journalists (-29), the news media (-27), and, especially, politicians (-55) divide us than say the same about any of the platforms we asked about.

Findings relating to platform policy and governance:

• On average across eight countries, more than one third of respondents feel that areas such as misinformation (39%), generative AI misuse (38%), and tech companies’ handling of personal data (36%) are getting too little attention from policymakers. Those in Brazil are more likely to think this, whereas in Japan people are more likely to say the amount of attention these issues get is about right.
• A majority in each country believe that platforms should take responsibility for showing false or misleading information posted by their users, with especially high support in South Korea (77% for social media and 79% for video networks) and Japan (73% for both social media and video networks), and a slightly smaller majority in the USA (65%) saying the same for social media platforms.
• Generally, in line with previous research by the Knight Foundation in the USA, there is much greater public appetite for platforms taking responsibility for their products and services than there is for government action. Generative AI is the main exception to this – for this platform, only 48% believe platforms should hold primary responsibility, while 39% think national governments should have a greater role.
• Views are split on whether platforms should only allow strong privacy settings versus letting users choose their own settings, with little variation by platform.

Findings relating to the public perception of platforms more generally:

• Despite the generally low trust and widespread concern over various issues related to politics, large numbers of respondents judge that platforms have a net positive impact both for them personally and for society as a whole. This is especially true for search engines, and, to a lesser extent, video networks. People in Germany, the UK, the USA, and South Korea think that social media has a net negative effect on society (while often thinking it has a net positive effect on them personally).
• While people use digital platforms, they are also concerned about problems associated with them, with strong majorities expressing concern over various online issues. The spread of misinformation online tops the list, with 87% expressing concern, followed closely by the use of generative AI to create fake content (84%). Concerns about personal data usage by major tech companies also remain high at 80%. By contrast, issues around freedom of expression see slightly lower levels of concern.

Introduction

Are digital platforms doing more harm than good to our democracies? And what – if anything – should policymakers do about them? These were the organising questions for a 2022 report on media and democracy published by the Knight Foundation and produced in collaboration with Gallup and a team of researchers based on survey data from the United States. In this report, we ask a similar set of questions, providing both a 2024 update on many of the issues covered by that report for the USA and, importantly, a more international look at how the public in different countries see and think about the different platforms that, in practice, constitute much of the digital public square.

If one looks to public opinion for answers on the question of whether digital platforms help or hurt democracy, the overall response is both. We document widespread public awareness of, and concern over, a number of issues associated with digital platforms, including the spread of misinformation, the use of AI to create fake content, and problematic data collection practices. We also show that many see platforms as having made problematic behaviours, including harassing people and spreading extreme viewpoints, easier. At the same time, however, the bulk of the public use platforms regularly for many purposes, including accessing news and information about politics, and see platforms as having made it easier to connect with people and find information. More people feel that platforms, for all their shortcomings, have made their personal lives better and bring societal benefits than say the opposite. Much of the public clearly recognise the complex downsides and upsides that many scholars also argue platforms offer. In short, we find evidence of ‘platform ambivalence’.

Young people are generally more optimistic about the role of platforms, even as they are often also keenly attuned to the problems at hand. Most people continue to be intense users, even as they worry about some of the implications. Many of these perceptions are broadly aligned with those the Knight Foundation documented two years ago, and our eight-country comparison documents that many of them are relatively consistent across different parts of the world.

Going beyond the 2022 report, we offer a more differentiated view here of people’s judgement of different kinds of digital platforms. Rather than asking about ‘internet technologies’ at large, we have a detailed battery of questions across different digital platforms, including public perception of their role in bringing people together or driving us apart. Here, two findings are particularly important to underline. First, public perception of social media platforms is much less positive than public perception of many other digital platforms. Much of what has been described by some pundits as a ‘techlash’ is, in reality, probably primarily a reaction to social media and problems ascribed to them, perhaps even to the real and perceived failing of specific companies. Second, while social media is seen as being among the factors that drive us apart, when compared with other social institutions, even more people say that journalists, news media, and, especially, politicians divide us.

What do people want, going forward? Looking across the eight countries, large minorities of respondents say that threats such as misinformation (39%), generative AI misuse (38%), and tech companies’ handling of personal data (36%) are getting too little attention from policymakers. It is clear that much of the public want more aggressive scrutiny of the role for-profit companies play in these areas. At the same time, in line with what the Knight Foundation found in the USA in 2022, across the eight countries we look at, there is generally much greater public appetite for platforms taking responsibility for their products and services than there is for government action. A demand for more policy attention and more scrutiny is not always the same as demand for a greater role for active, direct government intervention.

What people do with, say about, and think of digital platforms is not the only factor to consider as we seek to govern the digital public square, but it is an important factor to consider for anyone interested in developing forms of governance that combine credibility and efficiency. In this report, we provide an international and up-to-date basis of evidence that we hope can inform these discussions going forward.

Methodology

The data for this study were collected through a survey conducted by Ipsos, commissioned by the Reuters Institute for the Study of Journalism (RISJ) at the University of Oxford. The study’s main objective is to investigate public attitudes towards regulating the digital public sphere and to examine how individuals in different nations view and interact with digital platforms. This report builds upon earlier research jointly developed by Gallup, the Knight Foundation, and two researchers (Knight Foundation 2022). This earlier report was limited to the USA so here we broadened the analysis to include eight countries.

Ipsos collected data using an online questionnaire fielded between 1st and 25th October 2024 in eight countries: Argentina, Brazil, Germany, Japan, South Korea, Spain, the UK, and the USA. The country sample provides a range of different high-income democracies that previous research has documented have all embraced digital media and platforms, but have done so in quite different ways (Newman et al. 2024; Nielsen and Fletcher 2023). Japan and South Korea, for example, are important examples of countries where domestic platform companies are in many ways more prominent than US technology companies.

Ipsos conducted the fieldwork, handled data weighting, and provided data tables. Meanwhile, the authors (one of whom contributed to the above-mentioned Knight Foundation report) designed the questionnaire and managed the reporting and interpretation of the results. Each country’s sample was structured using nationally representative quotas for age, gender, region, and political orientation. The data were weighted according to census data or industry-standard targets for these demographic variables.

Each country’s sample size was approximately 2,000 participants. Due to the non-probability sampling method, a traditional ‘margin of error’ for individual data points cannot be calculated. We caution readers that differences of two percentage points or less are unlikely to be statistically significant and should be interpreted carefully because such small variations are generally not considered meaningful and are not emphasised in the analysis to prevent misinterpretation.

It is important to acknowledge that online surveys may under-represent individuals who are not digitally connected, such as older adults, less affluent groups, and those with limited formal education. Additionally, online panel participants often include a higher proportion of well-educated, socially, and politically active individuals who voluntarily opt into surveys.

Furthermore, online surveys present certain limitations, mainly due to their reliance on respondents’ memory, which is often susceptible to inaccuracies and various biases. Several sections of the survey ask respondents to recall past behaviours, beliefs, and attitudes towards various digital platforms – a task that may be influenced by social desirability bias.

Additionally, when addressing complex socio-technical issues, respondents might interpret terminologies differently from experts, potentially impacting the consistency of responses. While such challenges are typical in surveys like this, we have implemented strategies to mitigate these effects, including careful question design and wording.

Lastly, it is worth noting that data collection occurred between 1st and 25th October 2024, coinciding with election campaigns in the USA and Japan. While we do not observe clear implications of this for our findings, heightened political engagement during this period could have influenced views on political participation in these countries, which readers should consider when interpreting the results.

Chapter 1: Public use of platforms for political news and information

For more than a decade now, the impact of platforms like social media, search engines, video networks, and messaging apps on democracy has been a central part of the broader political discourse. And based on events in the last two years – following the emergence and rapid adoption of ChatGPT – we can add generative AI to the list of platforms that are seen as potentially shaping the public’s relationship with politics.

Perhaps the most basic question we can ask, then, is how widely are online platforms used for news and information about politics – and how do they compare with other sources of information?

The results from our survey show that search engines such as Google, Bing, and Yahoo! – used by 45% for news and information about politics in the last week – are the most widely used online platform type on average across the eight countries studied (Figure 1). However, they are less widely used for politics than both television (59%) and online news websites and apps (57%) – which we defined for respondents as being either the website or app of a broadcaster, newspaper, magazine, or digital-born source. Social media (41%) – which includes networks such as Facebook, X (formerly Twitter), Instagram, and TikTok – is the next most widely used platform type, followed by video networks (30%) such as YouTube, Vimeo, and Dailymotion. Messaging apps like WhatsApp, WeChat, and Facebook Messenger (15%) are less widely used for news and information about politics. The use of generative AI chatbots like ChatGPT, Google Gemini, and Grok (7%) for this purpose is still marginal, but this may change as they move from ‘visible’ standalone tools to ‘invisible’ features integrated into existing services (such as AI overviews in Google search).

Figure 1.

Figure 2 shows that some countries deviate from this general pattern. In the UK and the USA, search and social media are equally widely used for news and information about politics, and in Brazil and Argentina social media is ahead – in part because social media is more widely used generally. In Japan and South Korea, where parts of the platform ecosystem look very different, social media is relatively marginal when it comes to politics (as are messaging apps), but video networks and search are both widely used. It is important to recognise that these variations are not driven by whether people use digital platforms generally – social media, for example, is used by virtually all internet users in all these countries – but are variations in whether and how people use them to access news and information about politics (Nielsen and Fletcher 2023).

Figure 2.

In the Reuters Institute Digital News Report (Newman et al. 2024), we often refer to Japan and South Korea as ‘deeply aggregated’ online news media environments, because the public there are much more likely to rely on search engines and news aggregators as a gateway to news. In Japan, Yahoo! News is by far the most widely used online news source, and Line is a popular messaging app. In South Korea, a handful of domestic platforms structure the online media environment and are themselves important gateways to news. Departing from the Meta (Facebook, Instagram, and WhatsApp) and Alphabet (Google and YouTube) domination that we see in most other countries in the world, the most widely used social network and messaging app is KakaoTalk, and Naver and to a lesser extent Daum compete with Google for search market share. YouTube, however, is the dominant video network in South Korea – in common with much of the rest of the world. Nonetheless, these differences at the platform brand level probably, to some degree, shape the differences we see between South Korea and some of the other countries in our survey.

If we look at that usage data in a different way, and group together all online platforms (search engines, social media, video networks, messaging apps, and generative AI) and group together all online and offline sources of news (TV, print, radio, and online news websites/apps) we can see the overlap between each.

When we do this in Figure 3, we see that in all eight countries, most people – between 52% and 82% depending on the country – are now getting their political news and information from platforms and online and offline news outlets. Most people, then, are navigating a media environment that contains lots of very different sources of political news and information, and platforms – particularly search and social – have become deeply integrated into how they do so.

This is even true in the UK and Germany, which in many respects have more traditional patterns of news consumption, where only around one third in 2024 get news and information about politics exclusively from online or offline news outlets without also relying on online platforms. Elsewhere, the figures are even lower.

By the same token, very few people are only getting political news and information from online platforms. The highest figure in our data was 10% in Brazil. The same is true for younger people – just 10% of 18–24s get their political news and information solely from platforms. And even for these people, much of the news content they see on platforms will have originated from news outlets and will be branded as such.

Figure 3.

It is difficult to unpick the relationship between the use of online platforms for political information and the use of websites and apps of news outlets – because most platforms host information from these sources, as well as acting as gateways to them. Although platforms are used by some as a one-stop-shop for news, research based on passive web tracking data shows that on average they increase people’s use of news websites, as people often click through to articles they would not have accessed otherwise (Stier et al. 2021). We cannot do this kind of analysis with the survey data we have, but it is worth pointing out that people who use platforms for news and information about politics are more likely to also use online news from broadcast and print brands (Figure 4). For example, among those who do not use social media for news and information about politics, 22% use a newspaper website – but this rises to 39% among those that do use social media.

Figure 4. 

Although the overall picture is one where most people are using both online/offline news sources and online platforms, there are still relevant differences by demographic groups. Across all countries, those more interested in politics are more likely to use platforms for news and information about politics. Personal motivation for consuming news, which in media research is usually understood and measured as political interest, has long been seen as the primary reason why some people consume news while others do not. However, we might question whether political interest is as important in today’s platformed online media environment, given that people often see news and information about politics ‘incidentally’, as they use platforms for other reasons (Schäfer 2023; Tewksbury et al. 2001). But in fact, there are still very large differences in the use of all platforms for news and information about politics by political interest.

As Figure 5 shows, the differences by political interest are large for all five platform types, but particularly for search engines (26% not interested versus 52% interested), because they typically require the user to be motivated enough to enter a search query related to politics. The differences for social media – though still large (24% versus 47%) – may be slightly smaller because many social networks still enable a degree of incidental exposure to news and information about politics, whereby users are shown political content after logging on to do something else.

Figure 5. 

If we compare these gaps with those for offline sources – TV, radio, and newspapers – we see gaps of a similar size, though the largest gap of all is for online news websites and apps (34% versus 68%). Political interest matters both online and offline, and this is true for people in every age group. But it is also noticeable that TV is still relatively widely used by people with low interest in politics (40%), making it a particularly important source of political information for this group – even more so than social media.

One factor that is less often considered in the discussions about the use of online platforms for politics is how comfortable people are with using new technology – but there are big gaps here too (Figure 6). People who tend to think things like ‘new technologies make my quality of life worse’ or that they ‘find it difficult to keep up to date with new technologies’ are less likely to have used platforms for news and information about politics1. However, there are no such gaps when it comes to the use of offline sources of news and information about politics, such as TV, newspapers, and radio. This reminds us that there are still barriers to entry for some people when it comes to online platforms (even among people who participate in online surveys), and again, that TV is a particularly important source of information for this group.

Figure 6. 

There are also smaller – but still noteworthy – differences in the use of online platforms by age and gender as, Figure 7 shows. Again averaging across the eight countries, men are more likely than women to use all platform types for news and information about politics, and the difference is particularly stark when it comes to the use of video networks – used in the last week by 37% of men, compared with 23% of women. Younger people are more likely to say that they use each platform type for news and information about politics. The differences are particularly large for social media, which is used by 58% of 18–24s for news and information about politics, but by just 27% of the 55s and over. The differences by age group are smaller for the other platforms – and in the case of search engines, there is no significant difference between the youngest and the oldest age groups.

Figure 7.

One of the reasons that the differences by age are interesting is that they run counter to what we know about the use of other types of media for news and information about politics, which are much more heavily used by older people – even in the case of online news websites (Figure 8). If we return to the comparison with TV and radio – which are in a sense offline platforms, given that they offer people access to a range of different news and information choices in the same place – we see pronounced age differences, but skewed towards the older groups.

Figure 8. 

Older people are much more likely to get political news and information from television, and for the over 55s, it is clearly the most widely used source (69%). The data are also a reminder of the importance of television for the younger groups, with only social media more widely used for politics than television among the 18–24s. When it comes to gender, similar to platforms, men are more likely than women to use TV and radio for news and information about politics.

Finally in this chapter, we consider the differences in platform use by political leaning (see next page for Figure 9). Throughout this report we group respondents by whether they self-identify as left wing, right wing, or in the centre, because – although it is an imperfect measure, and there are other ways of grouping people – it is the only set of political categories that is widely understood by the public across different countries.

Figure 9. 

In terms of the use of different platforms for news and information about politics, we do not find large differences by political leaning in different countries (Figure 9). In most cases, the rank order of platforms within a particular country does not deviate from that described above. There are some exceptions. For example, those on the right in Japan are more likely to use video networks for news and information about politics relative to those on the left, and in the UK those on the left and the right are more likely to use social media than those in the centre – but in general usage patterns are quite similar. The bigger divide is between those who are interested in politics and those who are not. This is in line with a broader literature in the social sciences, which reminds us that while partisan divides are often very visible, because they are drawn between very vocal groups, the ‘other divide’, between those interested in politics and those who are not, is often at least as significant (Krupnikov and Ryan 2022).

Of course, this does not necessarily mean that the content people see on platforms does not skew in one direction or the other, or that people at different ends of the political spectrum do not think about platforms differently. We will explore this question later in the report.

Chapter 2: Platforms and their relationship to politics and democracy

Online platforms, including social media networks, search engines, video networks, messaging apps, and generative AI tools, have fundamentally transformed our media and information environment (Nielsen and Ganter 2022). Platforms are now understood by some experts not just as technologies for accessing news and information about politics (which we explored in Chapter 1), but also as having the potential to influence people’s relationship with politics and society (Fischer and Jarren 2024; Mont’Alverne et al. 2022).

Although some experts and pundits initially viewed platforms with optimism, concerns over privacy, algorithmic biases, and the spread of misinformation have led many to question their broader societal role (Metzler and Garcia 2024). This is especially true after contentious political events and when the stakes are high. Platforms are also, in many countries, increasingly explicitly politicised as elected officials, civil society groups, and rival industries question their conduct and role in society.

But what do the public think about the relationship between platforms, politics, and democracy? We first explore how trusted different online platforms are for news and information about politics. How trusted a platform is can be very different to how trustworthy a platform actually is, but the former is still important to understand because it tells us something about how people will likely respond to information they see there.

The results show that, on average across eight countries, search engines are the most trusted source of news and information about politics – with a slight majority of 55% saying they ‘somewhat’ or ‘strongly’ trust them (Figure 10). This relatively high level of trust in search engines stands in stark contrast to other platforms. Just over one third on average say they trust video networks (37%), and just under one third trust messaging apps (31%) – with a roughly similar proportion saying that they distrust them (29%). Social media is trusted by 30% and generative AI by 27% – and both of these platforms have higher proportions who say that they distrust them (38% for social media and 33% for generative AI). Generative AI tools, however, are less widely used by the public at present, and public perceptions may shift as people have more experience of using them.

Figure 10. 

Unpacking trust in platforms across different countries in Figure 11, we observe a relatively consistent pattern. Search engines are generally the most widely trusted platforms for news and political information, with video networks usually in second place. The relatively high trust in search engines might be viewed positively if we assume that they deliver trustworthy results, but it is important to keep in mind that search engines can return sources containing false information and research has found that ‘online searches to evaluate misinformation can increase its perceived veracity’ (Aslett et al. 2023). However, differences emerge with the less trusted platforms. For instance, in Japan and South Korea, messaging apps are met with notable scepticism, and are the least trusted platform for news and information about politics in South Korea (15%). Interestingly, in South Korea, generative AI (24%) is more widely trusted than social media (18%), but this is primarily because social media is viewed more sceptically here than in any other country in the survey (the next lowest is 24% in Germany).

Figure 11.

The breakdown of trust by age, gender, and political leaning in Figure 12 reveals distinct patterns. Search engines emerge as the most trusted source for news and information about politics across all age groups, but trust declines −9pp with age, from 62% among the 18–24 age group to just 53% for those 55 and older. However, search engines are still by far the most trusted platform in this older age group. The steepest declines by age are for video networks and social media, which start with higher trust levels among the 18–24 group (50% and 41%, respectively) but drop significantly among older adults, down to 29% for video networks and 22% for social media among the over 55s. This trend may reflect a generational gap in platform familiarity, with younger users typically having grown up with many of these platforms, alongside a preference among older individuals for traditional news sources, often seen as more reliable and consistent. Generative AI, as a newer technology, has low trust across all age groups, with a -17pp variation between the youngest and oldest groups – partly a result of its limited use as an information source (Fletcher and Nielsen 2024).

Figure 12.

Gender-based differences in platform trust are relatively small by comparison, though men generally express marginally higher trust in all platforms, particularly video networks (42% versus 33%). Political leaning shows a similar pattern, with those on both sides generally expressing the highest trust in search engines, while other platforms receive lower trust ratings. Trust is also higher on average among those on the right compared with those on the left.

Figure 13 shows that higher trust among those on the right compared with the left is true of most platforms in most countries. In the UK and Argentina, for example, every platform is more widely trusted by those on the right compared with those on the left. But there are some departures from this pattern. For example, search engines are more trusted by those on the left in Germany, Brazil, and the USA. And in some cases there are no meaningful differences between those on the right and the left – for example, with generative AI in Spain and the USA.

Figure 13.

While trust is often analysed through the lens of political attitudes or demographics, it can also be understood through broader views about society, such as people’s general optimism or pessimism about the direction they feel their country is heading in. At the aggregate level, we find clear variations between those with positive versus negative outlooks on the national situation (Figure 14).

Figure 14. 

Figure 14 illustrates how trust in political information from digital platforms varies based on this outlook. Those who think their country is heading in the right direction have significantly higher trust in each platform, with the ‘boost’ from this optimistic view being roughly the same for all platform types. This trend also suggests that a generally negative outlook may foster a broader scepticism around the information people consume.

We are also interested in what actions and behaviours related to politics and democracy people think platforms have enabled. We asked respondents about whether they think platforms in general have made certain things easier or harder. The results are displayed in Figure 15. Averaging across eight countries, clear majorities think that platforms have made it easier to connect with friends and family (66%), find the information they need (63%), and connect with likeminded people (59%). Despite sometimes intense elite and media debate focused on problems associated with digital platforms in the intervening years, these results are broadly similar to those documented by the Knight Foundation report in the USA in 2022, and cross-national findings from the Pew Research Centre² the same year. However, when it comes to the behaviours that are arguably most closely related to politics, only around one third think that platforms have made it easier to participate in politics (36%) or decide who to vote for (34%). But we should note that even in these cases the proportions are larger than those who say platforms make these harder – it is just that most people think platforms have made no difference. In fact, roughly similar proportions think platforms have made each of the actions we asked about harder (around 10–15%), with variation in the proportion who think platforms have made it neither easier nor harder.

Figure 15.

It is clear, then, that people recognise some of the benefits of platforms to them. But with a slightly different question about whether platforms have made it easier or harder for people in society to do a range of broadly negative things, people tend to think platforms have made these easier too. This is perhaps the clearest example of ‘platform ambivalence’ that we find in the data. More than half think that platforms have made it easier for people to say things they wouldn’t say in person (69%), spread misinformation (69%), harass or threaten others (66%), or spread extreme viewpoints (64%). Slightly fewer people think platforms have made it easier to interfere with elections (48%), but again, this is much larger than the proportion who think platforms have made it harder (12%). Many of these findings are aligned with what has been documented in the USA in the past by the Knight Foundation and the Pew Research Centre, and the patterns are broadly similar across the otherwise quite different countries we cover in this report. Most people, in every country we cover, are clearly aware of and concerned about many of the problems associated with digital platforms.

Of course, some of these actions only apply to certain platform types (search engines do not really allow you to freely express yourself, for example) – but the key point is that people do not generally think platforms are entirely good or bad. Rather, they recognise that it really depends on how people decide to make use of them.

The fact that people can use platforms to communicate in different ways raises the question of the effect they might have on social cohesion – whether they, on balance, divide us or bring us together. Of course, this is an empirical question and levels of social cohesion and polarisation can be measured. However, this is also something that citizens experience and perceive for themselves – which itself has consequences for how they behave.

On this basis we asked people whether they think platforms ‘divide us or bring us together’. Whereas the Knight Foundation report from 2022 we build on here asked respondents whether ‘the internet’ does more to divide us than to bring us together, here, to provide a more differentiated and nuanced view, (1) we split out the five different platform types and (2) to help us interpret the results, we compare them against other key institutions and groups in society. To simplify the results, we refer to the net difference between the proportion who say each ‘divide us’ and the proportion who say they ‘bring us together’.

Focusing on platforms first, we see quite large differences between each type in Figure 16. On balance, and averaging across the eight countries, people think that messaging apps (+19) and search engines (+12) bring us together. Messaging apps are primarily used for communication between friends, family, and other individuals people know personally, so this is perhaps unsurprising. It is harder to know for sure what is driving the positive results for search engines, but it could be that people think being able to quickly and efficiently access a shared body of information unites us in some sense, or it could be that search engines give us the information we need to be able to connect with others in person (finding local events, etc.).

Figure 16.

People think that, on balance, social media divides us (-13). Like messaging apps, social media can be used to stay in touch with people we know personally, but it can also connect us with people we do not want to connect with – making differences and disagreements more visible than they would otherwise be.

It is striking that, overall, people think both journalists (-29) and the news media (-27) not only divide us but are more widely seen as divisive than social media or any of the other four platform types that we asked about. This pattern is quite consistent across countries. The only exception is Germany, where social media (-38) is seen as more divisive than the news media (-18) – but even here people still see the news media as something that divides society rather than brings people together.

We also asked respondents about their perception of political bias on platforms. More specifically, we asked whether they think each platform is systematically biased towards a political view or not. On this issue, it is important to point out that many people (between 20% and 30%) say ‘Don’t know’, partly reflecting the fact that – as researchers in this area well know – it is very difficult to have a proper overview of what content platforms surface on platforms (something that may itself be seen as indicating the need for greater transparency).

Figure 17 shows that social media is the most likely to be seen as systematically biased, with 47% seeing it as such. This is greater than the 33% who do not think social media is biased in this way. Both video networks and generative AI are just as likely to be seen as biased versus unbiased, whereas search engines and messaging apps are less likely to be seen as biased. Only 34% view messaging apps as systematically biased towards certain political views – likely because the content people see there is less shaped by decisions made by algorithms designed by the companies themselves.

Figure 17.

On balance, people in different countries tend towards certain views (Figure 18). In Brazil, Spain, South Korea, and the USA, more people think platforms are systematically biased than not – given that the net proportion of those who say biased is larger than the proportion who say not biased. In Brazil, all platforms are seen as systematically biased, but in the other three countries it’s all platforms except messaging apps. In Japan, Germany, and Argentina, none of the platforms are seen as systematically biased, with the exception of social media in Argentina (+13).

Figure 18.

We can also break down the responses to these questions by political leaning. If we average across countries, the data suggest a surprising convergence across political lines. However, if we look at the left/right splits within countries in Figure 19, we start to see large differences in some cases. If we take social media as an example, people on the right in the USA are more likely to think social media is systematically biased towards certain political views than those on the left (69% versus 52%), with the opposite true in Argentina (48% versus 71%).

Figure 19. 

Despite the fact that less than half think that platforms are biased, when we asked people whether they are concerned that information on each platform is systematically biased towards certain political views, most said that they were either very or somewhat concerned (Figure 20). Repeating the pattern we saw in the previous chart, people are more likely to be concerned about the information on social media, and less likely to be concerned about the information on messaging apps. Despite a well-documented ‘third person effect’ in some areas, whereby people tend to think that everyone else is influenced by the media but not themselves, we do not find different levels of concern when we asked about the ‘information you see’ and the ‘information other people see’. These patterns are repeated across all eight countries in the survey.

Figure 20.

This pattern is replicated across all eight countries when we look at them individually, though levels of concern are generally higher in Argentina, Brazil, South Korea, and Spain. There are also no real differences by age and gender.

Chapter 3: Platform policy, responsibility, and governance

Over the past two decades, expert thinking around the internet and digital platforms has shifted dramatically. Once envisioned as a borderless cyberspace beyond national regulation, these platforms are now subject to increasing oversight (Flew and Wilding 2021). The rapid growth of platforms has raised concerns about their governance, responsibility, and impact on the public interest, prompting liberal democracies to craft new regulatory frameworks addressing disinformation, hate speech, and other issues.

While digital platforms empower individuals to create public value, they can also foster harmful content, polarisation, and ideological conflict, challenging democratic norms and public values (UNESCO 2023). Many scholars argue that tackling these challenges requires moving from a single-actor responsibility model to a collaborative governance approach involving platforms, users, and public institutions (Helberger et al. 2018).

Given the public’s crucial role as users, this chapter explores perceptions of platform policy and governance, examining whether the public view governance as a shared responsibility, their attitudes towards data collection, and their opinions on accountability for misinformation – key elements in fostering collaborative and responsive platform regulation.

To assess whether public concerns about digital platforms, highlighted in Figure 21, align with the attention policymakers give to these issues, we asked respondents if they believe the policymakers in the respective countries are paying enough attention to digital platform-related concerns. Figure 21 illustrates that, on average across eight countries, more than one third of respondents feel that areas such as misinformation (39%), generative AI misuse (38%), and tech companies’ handling of personal data (36%) are getting too little attention from policymakers. Notably, there are significant country-level differences, with over 40% of respondents in Brazil – where respondents are most likely to think too little attention is being paid – saying their policymakers pay too little attention to various platform-related issues, compared with under a quarter in Japan, where for many issues people think the amount of attention these issues get is ‘about right’. These findings highlight diverse expectations for regulatory focus, particularly concerning misinformation and data privacy.

Figure 21. 

However, public opinion regarding policymakers’ attention appears to be the function of, among other things, how much they trust politics in their respective country. Figure 22 shows that, on average across eight countries, individuals who ‘strongly’ or ‘somewhat’ distrust politics in their country are more likely to believe policymakers are not addressing key platform-related issues sufficiently. For instance, 47% of respondents who distrust politics feel that issues like misinformation and generative AI misuse are overlooked by policymakers, compared with only 37% and 34% respectively among those who trust politics. This trend extends across other issues, such as data privacy and the influence of tech companies, indicating a correlation between political trust and expectations for regulatory attention on digital platform issues.

Figure 22.

Figure 23 builds on this by asking respondents who they think should set content policies – platforms or national governments – across different types of digital platforms. While a clear majority favour platforms themselves taking on this responsibility for social media, search engines, video networks, and messaging apps (with 61–63% in support), there is a notable shift in opinion when it comes to generative AI. Here, only 48% believe platforms should hold primary responsibility, while 39% think national governments should have a greater role. This could be because people generally have a more negative view of generative AI – as we have seen at various points in this report – but it could also be that people see a greater role for government intervention when it comes to new and emerging technologies.

Figure 23.

The public generally preferring most kinds of platforms – despite the problems associated with them – to take the greater responsibility for policies or guidelines, rather than the government doing so, is in line with what the Knight Foundation found in its 2022 report in the USA. But, as the Reuters Institute Digital News Report (Newman et al. 2024) has previously highlighted, there are meaningful differences in public opinion between Europe and the USA regarding the role of government in regulating digital platforms. Here, we disaggregate the data to examine preferences for regulatory responsibility on either side of the Atlantic. Figure 24 shows that there is a general preference for platforms to self-regulate rather than relying on government intervention in both Europe and the USA, particularly for social media, search engines, and video networks. However, generative AI stands out as an exception to this trend. In Germany, the USA, and the UK, opinions are more divided, with a significant portion of respondents favouring government oversight (ranging from 43% to 51%) over self-regulation. In Spain, a slight majority still support platform-led regulation even for generative AI, but with government oversight receiving stronger support than for other platforms. This variation in opinion underscores the complexities of regulating digital platforms and highlights a potential divergence in governance preferences between regions.

Figure 24.

Given the heightened concern over the specific role of social media in misinformation and manipulation – underscored by the frequent appearances of tech CEOs in Congressional hearings on accountability and platform responsibility³, and increasingly intense political scrutiny of individual companies and sometimes the whole sector in top political circles in Brazil, the European Union, and elsewhere – we also explore variation in respondents’ views by political leaning. Figure 25 shows that, on average, a slightly larger proportion of right-leaning respondents (66%) support self-regulation by social media platforms compared with their left-leaning counterparts (59%). In Europe, this trend is particularly evident in the UK, where there is a 13pp difference between right-leaning (57%) and left-leaning (44%) individuals. Similarly, in Latin America, Argentina shows a 12pp gap (77% right versus 65% left), and Brazil a 13pp difference (74% right versus 61% left). Conversely, Japan and South Korea exhibit minimal differences between political groups. These findings align with the broader tendency of right-leaning individuals to support less government intervention and favour self-regulation (or, in some cases, individual responsibility), reflecting their general preference for limited state involvement in various domains (Janoff-Bulman 2023; Oreskes and Conway 2022).

Figure 25.

We next explore respondents’ views on identity verification across various digital platforms. In other words, should people have to provide proof of their identity when registering to use different platforms? Figure 26 shows that for most platforms, people do indeed think that users should have to provide proof. Most people support identity verification for social media (69%), messaging apps (65%), video networks (61%), and generative AI (61%), perhaps reflecting the fact that 69% are concerned that people say things online that they would not say in person (see Chapter 3), but also limited concern about potential misuse of personal data or risks of online data theft. However, opinions are divided on search engines, with 48% in favour and 41% opposed, suggesting that people perceive search engines as purely functional tools for accessing information rather than as platforms that require personal identity disclosure.

Figure 26.

Support for the idea of proving one’s identity is consistently lower in Japan, where fewer than half typically think that users should have to do this – though this is still a more common view than the opposite. And when it comes to using search engines in Japan, a clear majority think that people should not have to prove they are who they say they are in order to use them.

Figure 27.

Continuing the exploration of public attitudes towards platform governance, we now turn to preferences on privacy control. The results in Figure 28 indicate a split opinion. For each platform, a substantial portion of respondents favour platforms enforcing only strong settings, with figures ranging from 43% for search engines to 48% for generative AI. But at the same time, an almost equal proportion believe users should have autonomy to set their own privacy preferences. This near-even divide reflects ongoing public ambivalence about who should hold the reins over privacy, and the fact that this is one of the few issues where people do not have different views by platform suggests that it is not something that is front of mind for many people. There is some country variation beneath the averages, however. Across all platforms, respondents in South Korea are more likely to favour only strong privacy options being offered to users, with respondents in Japan, Argentina, and the USA usually favouring the idea of putting privacy in the hands of users.

Figure 28. 

Finally, because the overwhelming majority expressed their concern about false and misleading information (see Chapter 2), we asked respondents about who they believe should be held responsible. Figure 29 shows that, overall, a majority in each country believe that platforms should take responsibility, with especially high support in South Korea (77% for social media and 79% for video networks) and Japan (73% for both social media and video networks). In the UK and the USA, views are also strong but show slightly more ambivalence. In the UK, for instance, 75% think social media should be responsible, while 16% disagree and 9% are undecided. Similarly, in the USA, 65% support responsibility for social media platforms, with a higher proportion (24%) believing that they should not be held accountable.

Figure 29. 

The pattern remains consistent across most countries, with video networks generally receiving similar levels of responsibility expectations as social media, though messaging apps see more variability. This consensus highlights a public expectation for platforms to actively manage misinformation, especially in Asian countries, which show the strongest levels of support. It also reflects some variability in attitudes towards responsibility, with developed countries like the USA showing more divided opinions.

Chapter 4: Public perception of platforms more generally

This final chapter focuses on what the public think about platforms. In Chapter 2 we explored how people see the effect of platforms on democracy specifically, but here we focus on what people think of platforms in a more general sense (though, of course, for some their perceived impact upon democracy may shape their overall view.) More specifically, we look at whether people think different platforms have a positive or negative effect on individuals and society, and how concerned people are about some of the problems associated with the contemporary information environment (such as misinformation, the power of technology companies, and freedom of expression).

We start by exploring respondents’ overall perceptions of the impact of online platforms, both personally and on society. This essentially asks respondents to weigh up the different advantages and disadvantages of platforms in these two areas. When it comes to impact upon them personally, some of these will probably have nothing to do with politics and democracy. But when considering the impact on society, respondents are perhaps more likely to be thinking in democratic terms, alongside other factors.

We analyse the data by looking at the net difference between the proportion who say each platform has a ‘somewhat positive’ or ‘very positive’ effect on them or society and the proportion who say ‘somewhat negative’ or ‘very negative’. When we do this, we see quite large differences by platform and by country, but in most cases we find evidence that people see a net positive impact. Despite the reservations that people have when it comes to the use of platforms for news and information about politics, they clearly also recognise the wider benefits - reflecting what we call ‘platform ambivalence’.

Figure 30 shows that, in all eight countries, search engines are seen as having the most positive net impact, both personally and on society, of any platform. Even in Germany, where people tend to be more negative about the impact of platforms generally, large proportions think that search engines have a positive impact – much larger than the proportion who think the opposite (+47pp and +38pp).

Video networks and messaging apps are seen, on balance, as positive for people personally and for society – although this varies a lot by country. Messaging apps are viewed very positively in Argentina and Brazil, and as we have already seen, people in Japan and South Korea have a more negative view of messaging apps.

When it comes to social media and generative AI, we start to see examples where people, on balance, think there is a negative impact or views are evenly split. In Germany, people clearly think that social media has a negative effect on society (-27pp), with the same true to a lesser extent in the UK (-12pp), the USA (-8pp), and South Korea (-7pp). In Germany they also think that social media has a negative effect on them personally (-5pp) – though in the other seven countries people think social media is positive for them – especially in Argentina (+44pp) and Brazil (+38pp). People in Germany also think that generative AI is bad for society on balance (-9pp), and views are split in the UK (-) and the USA (+2pp).

Figure 30. 

Across the board, people tend to think that platforms have a more positive effect on them than on society, though the differences are not huge. While there is some evidence of a version of the ‘third person effect’ whereby people worry more about the impact on others than on themselves, in general people seem to assume the positives about platforms they experience are experienced by others, ultimately benefiting society.

If we look at the proportion who say they think each platform has a positive effect on society (instead of the net difference we used previously) by age and gender, distinct patterns emerge (Figure 31). Age, again, matters for positive evaluations of platforms, with younger groups generally expressing more favourable opinions. As age increases, positive perceptions for platforms like video networks, messaging apps, social media, and generative AI decline. Search engines, however, maintain high positive ratings across all age groups. When looking at gender differences, perceptions are often similar, with no substantial disparities for social media, messaging apps, and search engines. However, differences emerge for specific platforms. Men are slightly more likely than women to view video networks (+5pp) and generative AI (+8pp) positively. These findings highlight generational and subtle gendered divides, at least for some platforms, in how people evaluate the societal impact of technology.

Figure 31. 

This overall positive view of platform impact does not mean that people are not concerned about some of the societal issues that surround them. Quite the opposite, in fact. As Figure 32 shows, in every country surveyed, a clear majority are ‘somewhat’ or ‘very’ concerned over various issues, with over two thirds of respondents concerned across each category. The spread of misinformation online tops the list, with 87% expressing concern, followed closely by the related concern about the use of generative AI to create fake content (84%). Concerns about personal data usage by major tech companies are also high at 80%. By contrast, concern around issues to do with freedom of expression, though pronounced, are lower, highlighting that misinformation and data privacy are more likely to be front of mind for respondents. Concern is lower in some countries – most notably Japan, where concern over each issue is about 15pp lower than the eight-country average.

Figure 32.

Concern over these issues does not vary much by whether or not people use platforms for news and information about politics – though platform users tend to be slightly more concerned. Taking social media as an example, 83% of those who do not use social media for politics say they are very or somewhat concerned about ‘the spread of misinformation online’, but this rises to 90% among those that do use it. There may be something of a ceiling effect here, but it also highlights that people often continue to make use of platforms for news and information about politics even if they are concerned about some of the potential drawbacks.

Figure 33. 

Concern about these issues remains high across all groups in society, but we do see some differences by political ideology (Figure 33). Concerns around misinformation and AI-generated fake content are more widespread among left-leaning individuals, with 92% expressing concern – even as the extent of actual exposure to misinformation and its impact remain heavily debated in academic circles (Guess et al. 2019; Altay et al. 2023). In contrast, right-leaning individuals – although they are also highly concerned about misinformation – show heightened concern about freedom of expression. This is probably in part based on elite cues as a number of right-wing politicians, activists, and media have, mostly without providing evidence, claimed to be subject to censorship by platform companies. This pattern underscores not only ideological differences in prioritising misinformation versus freedom of expression but also the powerful role of media and political narratives in shaping these public concerns.

Conclusion

Based on survey data from eight countries, we have looked at how people use various different platforms, what they say about the role they play in their lives and in society, and what they want in terms of greater policy attention and responsibility for what happens online.

One overarching theme in the results is what we call ‘platform ambivalence’. Yes, many people use platforms for news and information about politics, but they also indicate that they are sceptical about the information they see there. Many people are clearly concerned about misinformation, bias, privacy, freedom of expression, and platform power, but they also recognise the benefits of easy access to information and staying connected with friends and family.

Use and appreciation goes hand in hand with widespread recognition of a range of real and perceived problems – reflecting ‘platform ambivalence’. More than half think that platforms have made it easier for people to say things they wouldn’t say in person, spread misinformation, harass or threaten others (66%), or spread extreme viewpoints. Social media, in particular, is also seen as doing more to drive people apart than bring them together – though even more people say the same about journalists, news media, and, in particular, politicians. The latter point is a reminder that the tech industry is not the only social institution experiencing something of a backlash from parts of the public today.

Misinformation, abuse of generative AI, and platform companies’ handling of personal data are areas that large minorities of respondents feel are getting too little attention from policymakers. But, while significant numbers of people want greater political scrutiny, even larger numbers think platforms should hold primary responsibility for dealing with such problems (rather than governments). This is similar to what the Knight Foundation found in the USA in 2022, and the general pattern holds both in the USA two years later and across the different countries we cover here.

Despite generally low trust and widespread concern over various issues related to politics, large numbers of respondents judge that platforms have a net positive impact both for them personally and for society as a whole. While there are countries in which some of the behaviours and attitudes analysed here differ across the political spectrum, they are generally not as partisan as one might assume, and more influenced by age, comfort with technology, and personal experience of using the platforms in question. Many of our respondents are, even if also intense users, sceptics towards digital platforms in many respects – but it is important to understand this scepticism in the context of a wider scepticism towards many other institutions in society, including the news media and the political institutions that one might hope could hold platform companies to account. Some may even see this as normatively desirable – that people can use platforms while remaining sceptical about them; that they can see the benefits of platforms while at the same time recognise concerns about their role in society.

As citizens, policymakers, and others consider what kind of digital public square they want in a world where much of the infrastructure we rely on for accessing information and expressing ourselves is owned and operated by private for-profit companies, we hope this report can provide evidence of the depth and complexity of the public’s experience of and thinking around the role platforms play – and might in the future play – in our societies. For that evolving digital public square to work, many citizens will expect it to be at least in part attuned and responsive to their preferences and practices, which we have documented here.

References

• Ada Lovelace Institute and The Alan Turing Institute. 2023. How Do People Feel About AI? A Nationally Representative Survey of Public Attitudes to Artificial Intelligence in Britain.
• Altay, S., Berriche, M., Acerbi, A. 2023. ‘Misinformation on Misinformation: Conceptual and Methodological Challenges’. Social Media + Society, 9(1), 20563051221150412. https://doi.org/10.1177/20563051221150412
• Aslett, K., Sanderson, Z., Godel, W., Persily, N., Nagler, J., Tucker, J. A. 2024. ‘Online searches to evaluate misinformation can increase its perceived veracity’. Nature, 625(7995), 548–556. https://doi.org/10.1038/s41586-023-06883-y
• Fischer, R., Jarren, O. 2024. ‘The Platformization of the Public Sphere and Its Challenge to Democracy’. Philosophy & Social Criticism, 50(1), 200–215. https://doi.org/10.1177/01914537231203535
• Fletcher, R., Nielsen, R. K. 2024. What Does the Public in Six Countries Think of Generative AI in News? Oxford: Reuters Institute for the Study of Journalism. https://doi.org/10.60625/RISJ-4ZB8-CG87
• Flew, T., Wilding, D. 2021. ‘The Turn to Regulation in Digital Communication: The ACCC’s Digital Platforms Inquiry and Australian Media Policy’. Media, Culture & Society, 43(1), 48–65. https://doi.org/10.1177/0163443720926044
• Guess, A., Nagler, J., Tucker, J. 2019. ‘Less Than You Think: Prevalence and Predictors of Fake News Dissemination on Facebook’. Science Advances, 5(1), eaau4586. https://doi.org/10.1126/sciadv.aau4586
• Helberger, N., Pierson, J., Poell, T. 2018. ‘Governing Online Platforms: From Contested to Cooperative Responsibility’. The Information Society, 34(1), 1–14. https://doi.org/10.1080/01972243.2017.1391913
• Janoff-Bulman, R. 2023. ‘The Myth of Limited Government’. Yale University Press, 28 April. (Accessed Nov. 2024).
• Knight Foundation. 2022. Media and Democracy: Unpacking America’s Complex Views on the Digital Public Square. Knight Foundation.
• Krupnikov, Y., Ryan, J. B. 2022. The Other Divide: Polarization and Disengagement in American Politics. Cambridge University Press.
• Metzler, H., Garcia, D. 2024. ‘Social Drivers and Algorithmic Mechanisms on Digital Media’. Perspectives on Psychological Science, 19(5), 735–748.
• Mont’Alverne, C., Badrinathan, S., Ross Arguedas, A., Toff, B., Fletcher, R., Nielsen, R. K. 2022. The Trust Gap: How and Why News on Digital Platforms is Viewed More Sceptically Versus News in General. Oxford: Reuters Institute for the Study of Journalism. https://doi.org/10.60625/RISJ-SKFK-H856
• Newman, N., Fletcher, R., Robertson, C. T., Ross Arguedas, A., Nielsen, R. K. 2024. Digital News Report 2024. Oxford: Reuters Institute for the Study of Journalism.
• Nielsen, R. K., Fletcher, R. 2023. ‘Comparing the Platformization of News Media Systems: A Cross-country Analysis’. European Journal of Communication, 38(5), 484–499. https://doi.org/10.1177/02673231231189043
• Nielsen, R. K., Ganter, S. A. 2022. The Power of Platforms: Shaping Media and Society. Oxford University Press.
• Oreskes, N., Conway, E. M. 2022. ‘From Anti-Government to Anti-Science: Why Conservatives Have Turned Against Science’. Daedalus, 151(4), 98–123. https://doi.org/10.1162/daed_a_01946
• Schäfer, S. 2023. ‘Incidental News Exposure in a Digital Media Environment: A Scoping Review of Recent Research’. Annals of the International Communication Association, 47(2), 242–260. https://doi.org/10.1080/23808985.2023.2169953
• Stier, S., Mangold, F., Scharkow, M., Breuer, J. 2021. ‘Post Post-Broadcast Democracy? News Exposure in the Age of Online Intermediaries’. American Political Science Review, 1–7. https://doi.org/10.1017/S0003055421001222
• Tewksbury, D., Weaver, A. J., Maddex, B. D. 2001. ‘Accidentally Informed: Incidental News Exposure on the World Wide Web’. Journalism & Mass Communication Quarterly, 78(3), 533–554. https://doi.org/10.1177/107769900107800309
• UNESCO. 2023. Guidelines for the Governance of Digital Platforms. UNESCO. (Accessed Nov. 2024).