The most relevant alternatives to established commercial platforms belong to the so-called Fediverse. This is a network of platforms of different kinds that share being open-source, decentralized, and federated. For example, Mastodon is similar to Facebook, Pixelfed to Instagram, PeerTube offers capabilities like YouTube, Loops is an alternative to TikTok, and Lemmy follows Reddit’s concept. A key difference is that for each Fediverse platform type there are many instances rather than just a single one. A Lemmy instance is essentially a small, separate Reddit. Instances can be set up by users themselves, so administrative control—and with it power—is distributed. They are often focused on specific topics, regions, or groups, set their own rules, and are moderated by community members. Yet they are all connected in principle: using the W3C ActivityPub standard, Fediverse platforms automatically exchange posts in the background. Users normally have an account on an instance whose topic and rules suit them, but can interact seamlessly with users and posts on other instances—even across platform types; for example, Mastodon users can communicate with people on Pixelfed. If desired, users can opt out: by blocking users, hashtags, or similar, or collectively by an instance cutting its connection to another instance (“defederating”) when the latter’s content or rules fundamentally conflict. Users can also move their account to another instance fairly easily, because the system is designed for that. This shifts much more power to users and creates a self-regulating ecosystem of instances: unlike centralized commercial platforms, users can simply change instances if, for example, new rules or privacy policies no longer suit them.

The federation model therefore offers potential for resilience, more balanced power, and organic evolution of features and rules tuned to user needs. The Fediverse does not automatically solve other issues—such as how to handle AI-generated content and disinformation. Privacy-compliance is also more complex: it is harder to verify how individual instances use data received via ActivityPub. Malicious instances might ignore deletion requests or alter information. Defederation is only a limited remedy for such problems.

Each instance decides whether to use recommendation algorithms. Many instances currently avoid personalized, algorithmic recommendations entirely. Many Fediverse users say they turned to it to escape the downsides of common recommendation systems. At the same time, broader mainstream adoption may require offering recommendation options on demand, since platforms without them can appear unattractive by comparison—ideally following principles of configurability and transparency to avoid repeating known mistakes.

In recent years the Fediverse has grown significantly in users and importance. Commercial platforms like Bluesky and Threads have joined the Fediverse. Still, it remains a niche with a very small market share. Most of the Fediverse is run noncommercially and developed under open-source licenses by volunteers for the public good. That is a major strength—there is no pressure to maximize profit against users’ interests—but it also creates market disadvantages: critics say the Fediverse needs better marketing to reach more people. Reaching more potential users is one issue; convincing them and helping with the switch is another: onboarding is often criticized as too complicated for less tech-savvy people. How to simplify and unify user experience without losing the system’s benefits, and what an effective public outreach strategy would look like, is still open.

The main user criticism of the Fediverse is simple: compared with established platforms, the familiar, relevant content is often missing. Because of copyright-based restrictions against unofficial tools that could import content from other platforms into the Fediverse, large-scale uptake is difficult. It is ironic that manual and automated reposting between big commercial platforms is now common, tolerated, or even expected—and that some large tech firms systematically commit mass copyright use to train AI models, scraping text, images, and content from almost the entire web. Interestingly, despite strong indications or even evidence, there are relatively few criminal consequences compared with the scale of the problem. For these and other reasons, large-scale content availability in the Fediverse remains limited.

Which strategies can help rising Fediverse platforms gain meaningful user numbers despite these barriers is still an open question. The “ethical alternative” image helps with certain groups but alone is not enough for most users. One approach is to focus on specific communities and media creators as pioneers by offering them real, tailored value. The Fediverse’s built-in possibilities for users to help shape platforms—features, rules, and structure—makes this relatively easy and could be a competitive advantage. As a complementary tactic, platforms might encourage individual users to repost or link to content, which is less problematic from a copyright perspective (“guerrilla interoperability”). Corresponding incentive models should be evaluated.