The best time to leave corporate platforms is now, says Robert W. Gehl

The researcher advocates for non-commercial, community-run alternatives to Big Tech, and discusses ethics, consent and scalability in the Fediverse.

The October 2026 presidential election left an open wound: the suspicion — now under investigation by Brazil’s Superior Electoral Court and Supreme Federal Court — that Meta and X interfered in the electoral process to favor one of the candidates. Lula’s campaign accused the platforms of algorithmic manipulation, reach restriction and biased use of artificial intelligence; Meta denied it, but its own Oversight Board admitted failures in election protection. Amid this climate of distrust, interest in alternatives to corporate networks is growing — and the Fediverse is emerging as one of them. It is in this context that Robert W. Gehl’s research becomes especially relevant. He is the author of Move Slowly and Build Bridges (Oxford, 2025), on community-run social media that challenges the power of Big Tech. Gehl currently researches alternative social media and manipulative communication as the Ontario Research Chair in Digital Governance for Social Justice at York University (Canada). He is also a member of the executive committee of the Association of Internet Researchers (AoIR) and an administrator of the aoir.social instance on Mastodon. In this interview, Gehl recounts his research trajectory — from readings in the political economy of communication to the Spanish Wikipedia editors’ strike —, discusses consent in research on decentralized networks, explains the Fediverse as a project of ethical covenant, assesses whether ethics or that covenant are scalable, and, at the end, comments on the Brazilian situation in the search for alternatives to Meta’s platforms.

You have developed a significant number of research projects and discussions on social media, platforms, and alternative social media. What motivated you to research and discuss the latter, and why would it be important today to deepen the research and debate on alternative social media?

It was quite a few things that influenced me to turn to alternative social media.

First, in the late 2000s, while I was doing the research that would result in my dissertation and my first book, I was reading the work of political economy of communication scholars. Dallas Smythe, Vincent Mosco, Tiziana Terranova, Robert McChesney. Those folks highlighted how value is produced in communicative systems, including how the audience is both consuming media messages but also producing themselves as machines to realize the value of commodities. Mosco and Terranova in particular alerted me to how “users” in digital media do a great deal of labor to make digital media work, from reading advertisements to contributing labor to content moderation.

In 2006, Google bought YouTube, and the lessons of the political economists were very apparent. Google paid $1.65 billion USD for YouTube, which at the time was an enormous sum. I immediately thought of all the YouTube users who had done the work of adding videos, commenting, and otherwise contributing – were they getting paid? No.

And Facebook was rapidly growing at this point, as well. I realized how the emotional and communicative work of Facebook’s users was the source of value in Facebook – it wasn’t really anything Facebook as a company was doing, beyond providing a framework for its users.

The final element that cemented it all was a labor strike at Wikipedia. Around the year 2000, Spanish language Wikipedia editors withheld their labor unless Wikipedia would convert to a nonprofit organization that was funded via donations and not by selling advertising space. This changed Wikipedia from wikipedia.com (its original URL) to wikipedia.org.

I wrote about all this and realized that if Wikipedia could do it, why couldn’t Facebook users? Or, if Facebook wouldn’t become a nonprofit (or better yet, a cooperative owned by its users), why can’t activists make their own, noncommercial, community-run social media?

And these questions are far more pressing today. The best time for us to leave corporate social media and have community-run social media was 2007. The second best time is now. And the good thing about now is that we have options, thanks to the hard work of activists and coders who have been developing alternative social media since 2010 or so. There’s the globe-spanning fediverse (think of Mastodon, for example) and ATProto projects like Blacksky to lead the way. It’s an exciting time to do this kind of research!

Given that this is a great moment to research alternative social media, you have already pointed out that it is not advisable or correct to study them the same way we study corporate social media. Why is that? What differences should researchers take into account?

The biggest difference is in terms of privacy and consent. For example, for many years, some of my colleagues in Internet Studies would gather large datasets from sites like Twitter to make sociological claims. The implication here was that Twitter data was not only public, but it’s in the public interest to find out what “the public” thinks about various matters. Let’s set aside whether or not one can make claims about “the public” based on Twitter data and instead focus on consent. Are Twitter users consenting to be research subjects? Are they consenting to have their data taken out of context to make claims? I think that if we answer “yes” to these questions, we’re participating in an extractive model eerily similar to that of surveillance capitalism.

Others can disagree with my Twitter example, but there are more cases of nonconsensual research that are clearly very bad. Consider the highly unethical manipulation of people by university researchers who partnered with Facebook in the “emotional contagion” study of 2015. Researchers from Facebook and their collaborators literally altered people’s posts in a large experiment to see if it would affect their emotions. That’s deeply reprehensible.

People who are using alternative social media to communicate are often aware of these practices. They are seeking out social media that’s not meant to develop profiles of them for profit, and they also think a great deal about the context of their communication. They do not necessarily want to be “the public.” I would strongly urge researchers to not approach data from, for example, Mastodon instances without seeking the explicit consent of the folks on those instances. At the very least, consult with the instance admins of each instance to see if such research might meet with community approval. And maybe rethink your methodologies – consider interview or participant observation methods over something like a big data-like approach.

Talking about instance admins, You are today one of the administrators of the aoir.social instance. How has this experience been? What has it brought to your perspective as a fediverse researcher?

It’s been quite rewarding, actually! I was very concerned the instance would not be well-received on the fediverse – after all, it’s run by the Association of Internet Researchers (AoIR). I was worried that folks on the fediverse would assume AoIR was there to study the fediverse in an unethical manner. But because the fediverse has a culture of codes of conduct and transparency about intentions, we could use our code of conduct document as a kind of diplomatic text to show our intentions. And now we are very, very well federated on the network. I still can’t believe I can meet someone from another part of the world – Brazil, for example! – follow them on the fedi, and learn more about their instance, all from our own social media that we control.

You argue, in one of your research projects, that the Fediverse is not only a technical achievement but a social one — a “covenantal fediverse” based on ethical norms that emerge from daily interactions. If the technical dimension can be copied (anyone can run an instance), but the ethical dimension cannot, what happens when the Fediverse grows beyond the communities that share these values? Is ethics scalable?

What I find most fascinating is that this collection of tens of thousands of autonomous servers and their communities can each develop their own codes of conduct that are unique to them, and yet they can still connect to others in the network. And, most importantly, they can break the connection to servers populated with people with whom they do not align ethically. The result of this is a bottom-up collection of ethical principles, rather than a top-down imposition of values from a corporate headquarters.

It’s less a question about the universality of some ethical code and more a matter of what Iris Marion Young calls a “thin set of global values” that allows folks to communicate with one another. It’s really quite remarkable. So if by “scale” you mean universality, I would say no, ethics doesn’t scale. But if by “scale” you mean ethical values emerging from autonomous groups and aligning via a particular form of communication, then yes, ethics can scale.

Are alternative social media, the fediverse, the solution to the scenario of extractive big tech platforms?

Yes, on the condition that the alternatives are noncommercial and run by their community members.

Finally, would you like to add something to our conversation?

I would add that I am very keen to learn more about alternative social media in Brazil right now. Brazil has long fascinated me ever since the days of Miguel Freitas’s Twister. Now I learn that folks are building out the fediverse because they are leaving Meta behind – and for good reason. I see similar movements in Canada, where I live. I want more of this – more folks taking their communication infrastructure out of the hands of corporations and putting it back in the hands of everyday people!

Rolar para cima