It’s that time of the year, and I’ve made my annual pilgrimage to the annual conference of the Association of Internet Researchers (AoIR), the single most important highlight of the academic year. This year we’re in Niterói, Rio de Janeiro, and after the local welcomes we start the conference proper with a keynote by the great Marie Santini from NetLab at the Federal University of Rio de Janeiro, who is also a genuine Niterói local. She begins by revisiting the timeline of Internet studies: we have now reached a moment of great rupture (the theme of this year’s conference), but this also emerges out of a much longer timeline, stretching from the digital utopia and great optimism of the mid-1990s, when the emerging Internet and Web first caused significant change.
The response at the time was somewhat naïve: we believed the Internet would promote diversity, democracy, open access, and creative autonomy – this was encapsulated most prominently in the libertarian Declaration of the Independence of Cyberspace, which saw networks as a free and unregulated territory, echoing the American Declaration of Independence. Ruptures were welcomed at the time, as a challenge to existing hegemonies – but this ignored ethical, economic, and existential issues in the pursuit of such total disruptures of the means of mass communication.Silivon Valley still represented hope.
Twenty years on, the situation was much different: mis- and disinformation had emerged as amongst the most serious risks to global democracy, leading also to the election of Donald Trump and other autocrats around the world. This was worsened by the impact of the global COVID-19 pandemic, and the role of disinformation in encouraging vaccine hesitancy; despite its leading role in vaccine development, the US ranked first in COVID-19 deaths, and Brazil, facing a similar flood of disinformation, ranked second – even in spite of their substantial differences. What they had in common was mis- and disinformation, also spread directly and deliberately by their illiberal heads of state. This resulted in a lack of effective COVID-19 responses in both countries.
Further on, 2024 was one of the most comprehensive election years around the world, with more than half the world’s population voting for their national governments – but many such votes went to illiberal and antidemocratic forces. Some of these elected leaders – especially the Trump administration in the US – also aligned themselves with the leaders of Big Tech companies, who aided their reshaping of communication environments.
But if we are in the middle of another long cycle of twenty years or so, there is still a chance for this to turn around, and become a virtuous cycle. But what needs to happen for this? Back in the day, the optimists were clearly too optimistic; but the pessimists were also wrong in assuming that nothing much would change. Our mistake was to ignore the business models of digital platforms, and use this to anticipate their actions; we are perhaps making the same mistakes again with the rise of artificial intelligence now. In social media platforms, data, profit, and opacity are directly connected by the business model; the platform is incentivised to become a black box. Today, disinformation has become an industry: the lie is a commodity, with market value, a production industry, and a normalisation of such false content – the disinformation industry is now more profitable than the cultural industries, but we are not reflecting on this enough. The normalisation of fake news has numbed our critical capacities, but we need them now more than ever.
We can learn lessons from Brazil here, though, which has played a leading role in addressing these issues and reclaiming its digital sovereignty; notably, it also held former President Jair Bolsonaro responsible for his disinformation-fuelled coup attempt. The country passed a Big Tech regulation bill in 2023, temporarily blocked X for failing to comply with regulations in 2024, passed an artificial intelligence bill in 2025, saw Supreme Court rulings against Big Tech about their responsibility for disinformation on their platforms, and has moved towards legislation on digital childhood.
This is also driven by critical research activity, of course, which has also drawn significant attacks from Big tech companies. Doing research is therefore also a form of resistance against the extractivist business models of US-based platform companies – a business model which has caused instability in the whole world. Such attacks have targeted researchers’ funding sources in both Global South and Global North, but Global South researchers are more used to this situation; while some US and European researchers also receive funding from platform operators, this has never been the case in the Global South, requiring researchers to develop innovative approaches to working around data access and other limitations.
To do science in Brazil is to make the impossible possible, therefore; public universities have limited and precarious funding, yet produce the lion’s share of the research. To survive here it is necessary to fight problems like disinformation, not just to examine them. This does not protect researchers from the powerful reach of Big Tech, however, and there is a significant need for researchers to network and collaborate in their activities.
Three key themes for research emerge here: the first is transparency. Academic research is critically important as a guide to policy-making, and this demands detailed investigation building on access to data; platforms tend to eschew such transparency, but we must continue to demand such access rather than just the theatre of transparency that platforms currently engage in. Only this enables us to test the transparency promises of the Big Tech companies, and Marie’s NetLab will soon launch a global transparency index assessing the platforms’ performance. This also needs to be supported by policy, building on and further extending models such as the European Union’s Digital Services Act.
A second key agenda is to better understand the business models of platforms, which today means especially also the business models of AI companies. There is a need to audit their operations, and to avoid toxic commercial behaviours that have deeply problematic social and societal consequences. What happens when AI companies manipulate their models out of commercial, political, or other underlying interests? What happens when the models we use everyday are inherently biased by such manipulation? What are the consequences for society and democracy? Further regulation to avoid such scenarios is urgently required here, and the platforms’ performance must be closely audited.
Finally, environmental and climate disinformation points to a third key research agenda. It is clear that such disinformation has a direct effect on public and societal action; and disinformation has now shifted from outright denial to discourses of delay and greenwashing. NetLab will tomorrow launch a report on greenwashing advertising which shows a very high volume of such content on social media platforms; this feeds on the intransparency of social media platforms and aids the interests of fossil fuel industries and other groups seeking to frustrate effective climate action. This also now includes AI companies, of course, with their exceptionally high energy requirements. Polluting companies are known to engage in coordinated disinformation campaigns to delay such climate action, and there is an urgent need for more independent research into such campaigns, especially also from the Global South, and more information to the public on such campaigns and their impacts.
But today media companies also depend on such platforms, and therefore fail to effectively hold them to account; this undermines society’s ability to act collectively. More than ever, science, academia, universities have a critical role to play in pursuing these aims. Without access to data, there is no research; without the free knowledge produced by such research, there is no democracy.
