To fix social media, we need to introduce digital socialism

Proprietary social media networks need to be transformed into local and global digital commons.


Michael Kwet
Michael Kwet is a Visiting Fellow of the Information Society Project at Yale Law School.
19 May 2020

Facebook is among the Big Tech companies dubbed the 'frightful five', along with Google, Amazon, Apple and Microsoft [File: Reuters/Dado Ruvic]


In the past few years, intellectuals across the spectrum have fallen out of love with Big Tech. The “frightful five” – Google, Amazon, Facebook, Apple, and Microsoft – wield enormous power, with their combined wealth exceeding $5 trillion. After years of market consolidation and exploitative practices, critics in the United States are pushing to break up Big Tech monopolies with antitrust law.

Social media networks are a centrepiece in this conversation: they violate privacy, amplify sensational content and “fake news”, and manipulate users to keep their attention.

Breaking up Big Social Media sounds great, but how would this look in practice? A number of legal scholars and politicians have proposed reforming social networks by using antitrust law and regulations to create a more competitive marketplace. Leaving this task to market forces, however, is a bad idea that will not solve the central problem: proprietary control of the networks and the exploitation of user attention for profit.

A solution based on digital socialism is needed to transform social media into a global democratic commons. This would eradicate Big Social Media by placing ownership and control directly into the hands of the people.
The antitrust proposal

Various scholars have put forward two main ideas to break up Big Social Media, neither of which can sufficiently accomplish their goals.

The first one seeks to dismantle past mergers and acquisitions. Facebook, for example, bought up Instagram and WhatsApp years ago, and is now seeking to integrate all three platforms into a seamless communications network.

Scholars like Tim Wu, Sarah Miller, and Matt Stoller have suggested breaking Facebook, Instagram, and WhatsApp into three separate companies. They hope these companies would then compete for customers, which would compel them to treat users with respect.

Yet there is no good reason to believe this would do much for privacy and competition itself does not necessarily curb harmful behaviour. Even if these companies are broken up, given that their business model is based on serving ads and the exploitation of user data, they would have no serious incentive to change their behaviour.

Furthermore, these companies are able to monetise surveillance because the data is running through their platforms, and they force people to be a part of their networks in order to interact with their friends and family. For example, a user who does not like Facebook’s privacy practices can leave for another network, but then they have to convince their friends to join them.

The second idea proposes a solution to this problem: make social networks interoperate. Social media platforms would be forced to allow members of one network to interact with members of another. For example, a Facebook user would be able to post a comment under a YouTube video while logged into Facebook, and vice-versa. Users’ data would also be “portable” so they could move their profile to a different platform.

Interoperability exists in other communications services, such as telephone networks and email.

However, the “competition through interoperability” antitrust proposal is deeply flawed.

The reason Big Social Media firms are able to raid everyone’s data and mistreat users is that they are centralised, cloud-based intermediaries. If I want to share a photo with you, I first upload it to, say, Facebook’s servers, and then you download it from Facebook’s servers. The user experience is then determined by Facebook’s network software.

This form of cloud-based centralisation gives corporations power over the platform and the data. Proprietary network ownership provides corporations with the coercive power to monetise user data and force ads on users. Making proprietary networks interoperable does not change this power dynamic. The companies will simply compete to collect more data and serve more ads so they can generate profits.

Some antitrust scholars have also suggested social media networks can solve this problem by charging people to use their networks. Users who do not like spying and ads can pay out of pocket for social networking instead.

A subscription-based social network might sound great for the middle and upper classes, but it is a non-solution for the billions living in poverty. Those with little or no income are not going to “pay for privacy” or any other “exploitation-free” benefits, such as ad-free access.

This same conundrum plagues the mobile app ecosystem, where 70 percent of apps spy on users through hidden trackers. Proprietary control of the apps prevents the public from stripping out the trackers, and competition among millions of apps does nothing to prevent app publishers from mistreating users.
Building a social networking commons

The new antitrust proposal will fail to remedy social media ills because it is wedded to competition in a capitalist system. A genuine solution must, therefore, eliminate the profit incentive and give people direct control over the means of computation.

To fix social media, activists and lawmakers need to press for digital socialism – a commons-based solution embodying libertarian socialist principles of self-governance, decentralisation, and federation. Social media would be transformed from a profit-seeking enterprise into a global democratic commons. A technological foundation has already been created (as I detail below).

To see this through, we need to pass laws imposing decentralised, free and open-source technology solutions on the social media ecosystem. Big Tech corporations would be forced to relinquish user data, and social networking infrastructure would be owned and controlled by the users. The platform software would be open source so the public can inspect the code and customise the user experience.

To ensure the network infrastructure will be well developed and maintained, governments would subsidise public interest technology. Technologists could be paid to develop software at public universities and non-profit organisations. Developers across the world would collaborate and borrow code from each other, while individuals and communities would join networks or form new ones as they see fit.

Governments could also subsidise the rollout of broadband internet and personal cloud infrastructure. Inexpensive FreedomBox devices could be provided to lower-income households and small server farms could be operated by local communities.

Corporations might participate in some form, but the new laws and technology would effectively cut off their ability to privatise control, generate large profits, push ads, or spy on users.

Funds for implementing social networks can be raised by taxing the rich and Big Tech firms. Resources for infrastructure and development should be extended to people in the Global South as compensation for colonialism, including recent revenue extraction through digital colonialism.
Social media decentralisation

The foundation for a commons-based social media system was laid in the establishment of the Fediverse – a set of interoperable social networks based on free and open-source software. Fediverse platforms include Mastodon (akin to Twitter), PeerTube (akin to YouTube), and PixelFed (akin to Instagram).

The Mastodon social network, which has more than four million registered users, is the most polished example to date. Its feature set resembles Twitter: you can post to your wall, “like” and “share” other posts, follow user accounts, and so on.

However, there are a few crucial differences.

For one, there is no central server or administrator through which all user activity, data, and membership flow. Instead, you join one of many servers, called “instances”, which host and transmit user data. Each instance sets its own terms and conditions: it might ban hate speech and pornography, or focus on a shared hobby or interest.

To open an account, you simply sign up with an instance. Let us say you pick the username Alice at an instance called instance123.social. Your social media handle would be: @alice@instance123.social. Alternatively, you can pay to host your own instance and set the code of conduct to your liking.

The Fediverse uses shared communications protocols like ActivityPub so that users can interact across platforms. For example, a user from Mastodon can post a comment or follow a user from the PeerTube social network without ever leaving Mastodon. This is similar to email, where you can send messages from a Gmail account to a Yahoo account.

With Twitter, you have one timeline that displays posts and activity from other users. With Mastodon, you can pick from three timelines. The first is your home timeline that displays content (such as wall posts or videos shared) by the people you follow. The second is a local timeline that displays content from members of your instance. The third is your federated timeline, which displays content from other instances. Each timeline provides a different way to interact and discover content.

To make sure the experience is safe and enjoyable, Mastodon builds in a variety of content moderation policies. Individual users can filter out other users and instances that they do not want to see or interact with. Instance moderators can also filter out other users or instances. For example, if another instance is loaded with white supremacists, then you or your instance administrator can block that instance.

The ability for individuals to create their own instances, interact across networks, and set their own code of conduct undermines the centralised ownership and control model of Big Social Media. And because the server software powering Mastodon falls under a strong Free Software licence, the public can modify it to make it work as they wish.

For example, some developers created Glitch, a modified version of Mastodon which has its own set of features built in. In Glitch, you can set your posts as local-only so that they will not show up in outside instances.

The open sourcing of the network software also creates direct accountability to the public. If the Mastodon developers tried to, say, place banner ads inside their platform, an outside developer could take the code, strip out the ads, and release an ad-free version to the community.

The current Fediverse model is mostly decentralised, but there is room for improvement. Server administrators still possess the authority to surveil users and impose content moderation decisions on instance members. This means users have to trust the server administrators they interact with. To address this feature, Free Software developers are creating peer-to-peer technologies that fully distribute power and privacy down to the end users.

The LibreSocial network offers a glimpse of how this can work. There is no need to trust server administrators because the peer-to-peer architecture eliminates them altogether. Instead, the social network is operated by the community of end users through the LibreSocial software. The network is free and open-source, easy to use, and allows for customisation of the user experience – such as how to visualise a user wall, or social games – through the use of plugins that anyone can create or download.

While LibreSocial is still in a testing phase (which will soon be open to the public), the developers have built an impressive model for a fully decentralised social network.
Replacing digital capitalism with digital socialism

The ingenuity of the free software community is central to the struggle for tech rights and equality. Solutions like the Fediverse and LibreSocial prove that a world in which users are not exploited is possible. But they alone cannot pull away the billions of users stuck inside Big Social Media.

Ultimately, activists will have to push for new technologies, laws, and regulations that eradicate Big Social Media and transition the world to a social media commons.

Unfortunately, current legislative proposals by US legal scholars and Congress promote a capitalist model where “many Facebooks and Twitters” compete to capture data and user attention. This will not solve our problems.

Just as we cannot fix the climate crisis with “clean” coal, “all of the above” energy solutions, or cap-and-trade market-based reforms, we cannot fix social media with corporate owners, proprietary technology, centralised clouds, and market competition. The master’s tools will never dismantle the master’s house.

A genuine solution requires digital socialism: a decentralised social media commons based on free and open-source technology, supported by laws and the public purse. The foundation is already set, but a popular movement is needed to see it through.

The views expressed in this article are the author’s own and do not necessarily reflect Al Jazeera’s editorial stance.


Michael Kwet is a Visiting Fellow of the Information Society Project at Yale Law School. He is the author of Digital colonialism: US empire and the new imperialism in the Global South, and hosts theTech Empire podcast. His work has been published at Motherboard, Wired, BBC World News Radio and Counterpunch. He received his PhD in Sociology from Rhodes University, South Africa.