Why open protocols matter now
Telling my readers that open protocols matter feels like preaching to the choir, so I mostly don't focus too much on it. However, I think this story is worth flagging because it is such a clear signal of what's to come for 2026:
The CBS News show 60 minutes was going to air a story about the men who have been deported from the US to the El Salvadoran concentration camp CECOT. This story was killed last minute by the new editor-in-chief Bari Weiss for blatant political reasons.
The story was however broadcast in Canada, which lead to it being recorded and distributed online.
The episode of around 12 minutes (confusing show branding tbh) is now available to be seen on Bluesky:
The full spiked 60 Minutes CECOT package, clean & subtitled. 1/5
CBS and its owner Paramount are trying to take the story down, filing takedown requests on copyright ground.
People on Bluesky were commenting (1, 2, 3) about how this relates to fascism, that news unfavorable to the government have now to be smuggled into the country and distributed in manner that is resistant to censorship, drawing historical parallels.
I think these comparisons are broadly correct, and also set up for what's to come for open social protocols. This will not be the last news story that the US government will try to take down, in collaboration with legacy media organisations.
This puts open protocols like atproto in a new front line: it can be used to distribute news and reports that are unfavourable to governments that are trying to prevent the news from getting out.
The technical implementation of atproto is fairly capable of resisting takedown pressures. The practical reality of virtually everyone using Bluesky to access the atproto network creates a massive singular focal point for governments to apply pressure to take down certain content however.
I'm expecting that we'll see more of these dynamics in 2026. Atproto provides a place for sharing news stories that are taken down on the Big Tech platforms. This usage will likely increase pressure on the Bluesky company by governments to take down unfavorable political content.
The contestation of spaces
Bluesky is not the only microblogging atproto platform anymore however, with at least four projects that are setting up their own atproto infrastructure to build an interoperable microblogging network. Blacksky is currently the furthest along with this, but other projects are looking to create their own spaces as well:
Gander just passed their crowdfunding target, raising over 1.7M CAD from over 2100 investors, to build a Canadian social platform on atproto.
Eurosky just announced that in January they will offer Eurosky PDS hosting, their first step for European infrastructure on atproto.Northsky is starting host the first group of people on their own PDS as well.
Having multiple platforms that are interoperable with each other on the same network leads to two challenges that I'm watching in 2026:
The geopolitical aspect of handling political content. Above I suggested that Bluesky might get pressured by the US government to remove political speech from their app. How will the dynamic on the network evolve if these posts can still be viewed on the other platforms, whether that's Blacksky or Gander?
Bluesky currently has a reputation for highly aggressive comments and replies around certain contested topics such as AI or the economy. I think this is a signal that people experience platforms like Bluesky as a digital place, and that they care that the prevailing opinion in this place aligns with their own personal viewpoints. For topics like AI, where opinions diverge massively and are highly contested, this leads to very aggressive posting and reactions, not only because people disagree with each other, but also because they are contesting the 'Bluesky' space on what will be the prevailing opinion regarding these topics.
My current thinking is that the introduction of multiple interoperable spaces on atproto will lead to even more conflict and aggressive posting on the network, especially for subjects around AI. Now it becomes important for people that the dominant opinion on Bluesky aligns with their own opinion, but that this happens on Northsky and Eurosky as well.
Because posts from a Northsky PDS are visible to Blacksky users, and vice versa, it means that the "place" being contested isnt a single app, but the shared network layer instead. Because all projects make the entire network layer visible, it means that a conflict over whether Bluesky is pro-AI or anti-AI also automatically is a conflict over whether Gander is pro-AI or anti-AI. The stakes become higher, as people can more strongly identify with a specific app/platform. But because that app or platform contains the access to the entire network, and only provide different windows to the same underlying conversations, there's no actual retreat. You can identify as a Gander user or a Eurosky user, but you can't escape into a separate discourse. This results in the emotional stakes of platform identity get combined with the inescapability of a shared network layer.
So what happens when the identity of an atproto platform and wider network culture move in opposite directions?
Meta-communication
Last week saw a lot of discourse in the community of atproto devs about how to handle (heh) the login screen for atproto-enabled apps, and how to communicate the login flow as clearly as possible to a wide audience. There are challenges regarding how to communicate to a wide audience what their domain/handle/username/internet handle is that they can put in the box, as well as how to communicate clearly that this is all based on an open network that's more than just Bluesky.
For an overview of the conversation itself, I think this article by Chris Shank is great:
He concludes:
Ultimately I'm not sure what we should call handles or what the best branding for atproto's login flow is. So let's make sure the right people are researching this and to include accessibility and non-english speakers in this process. It is clear to me, however, that a unspoken politics and potential symbol is emerging out of this conversation.
Chris gets to the core here: what is on one hand a fairly obscure discourse about UX design, also displays much of the underlying questions about politics and power that go with building new social platforms on open protocols.
I'm curious how the following will develop in 2026:
How will the developers of various atproto apps collaborate, compete and communicate on aspects of atproto and interoperability? So far this mainly happens via posting on Bluesky with largely in-crowd people, how will that dynamic evolve? Here it's also worth looking at the fediverse and ActivityPub for some lessons, where involvement with protocol development was highly structured and formally open to everyone, but in practice dependent on a tiny group of self-selecting people. What will become the infrastructure for atproto to have these conversations?
What will actually be a good way to communicate important information about atproto to regular people? There's quite a few aspects on accessibility to incorporate here, as well on language more broadly. On one hand, there is a universality in the claims and suggestions for login screens. This is especially noticeable with 'internet handle', suggesting your atproto identity for the entire internet. But at the same time, many of the suggestions are build around the @ symbol, and the wordplay in English with AT. This works well enough in English, but for a universal suggestion it needs to be able to be used a a global majority of people who don't speak English.
Connected Places
Finally some musings about my own work: you might have noticed I've been experimenting new ways to share information about the ecosystem. I've been building my own website to share links that are on-protocol:
For early next year I'll be continuing with the experimentation. For most of the year I've been focused on being complete, giving people a full overview of the ecosystem. But as developer activity around atproto is picking up, this becomes both less practical and less useful.
For 2026 I'll be focused more on the curation aspect, and experimenting with new ways you can use the affordances of atproto to curate and share information about the network.
Besides that, I'll continue to write regular longform analysis, as thats what I enjoy doing.
Thanks for reading this year, and for all your support, it is greatly appreciated!