Bluesky publishes proposals for new moderation and safety tooling, focusing on user lists and reply controls for community-driven moderation, hashtags, and more
The goal of Bluesky is to turn social media into a shared public commons. We don't want to own people's social graphs or communities. @bossett.bsky.social , @davidgerard.co.uk , @pfrazee.com , @tananarivedue.bsky.social , and @bsky.app Mastodon: @carnage4life@mas.to , @shoq@mastodon.social , and @lrhodes@merveilles.town Twitter: @keithdsouza and @catalinagoanta . Forums: Hacker News Bossett / @bossett.bsky.social : The moderation discussion is starting to roll with the proposals now out at https://github.com/... & wanted to re-contribute this. — There are people saying ‘focus on the tech, moderation later’ & I think that's coming from entirely the wrong place. [embedded post] @davidgerard.co.uk : the bluesky moderation proposal is fucking wild. — * you MUST label your child abuse photos. Not remove them or put you in fucking jail - *label* them. — * Labeling DMCAed content is mandatory. (Internationally.) — * Nazi content, specifically illegal in multiple countries? That's fine! … Paul Frazee / @pfrazee.com : hey everybody, we're dropping a bundle of proposals about moderation & safety (& also hashtags) today — we'd really like to hear feedback on them to make sure we're pursuing the right ideas and not missing anything important [embedded post] Tananarive Due / @tananarivedue.bsky.social : It's kind of weird for me to be excited about info that feels like a town hall meeting, but here we are — https://blueskyweb.xyz/... @bsky.app : We believe that a public commons is important for social media. These proposals for moderation and safety tooling have been in the works for a while, and we're excited to share them for community discussion and feedback with you now. — https://blueskyweb.xyz/... Mastodon: Dare Obasanjo / @carnage4life@mas.to : Took a look at the BlueSky moderation proposals and they lay the foundations for choose-your-own-moderation to go along with choose-your-own algorithms via custom feeds. — 1. Double down on lists of users. Don't just use them to create feeds. … @shoq@mastodon.social : Incoming! — https://github.com/... https://blueskyweb.xyz/... cc @atomicpoet @tchambers @mekkaokereke @ben @davidslifka @davetroy @rogermcnamee — @Green_Footballs L. Rhodes / @lrhodes@merveilles.town : A few scattered thoughts about #Bluesky's latest post on moderation: https://blueskyweb.xyz/... (brought to my attn by @Roundtrip ) — What stands out most is how they're positioning themselves as a corrective to corporate social media, despite a structure that actually retains the data collection practices of corporate companies and expands the use of algorithms. … Twitter: Keith / @keithdsouza : Lol, jack wants to be the new w3 for social. This is hilarious at best. https://twitter.com/... Catalina Goanta / @catalinagoanta : Bluesky published its content moderation vision. ‘The goal of Bluesky is to turn social media into a shared public commons’. Here we go again. If it's a *company* that keeps on developing *products* they *sell* to *businesses*, that is not a public commons https://blueskyweb.xyz/... Forums: Hacker News : Moderation in a public commons
Context & Ripple Effects
Bluesky has spent two years building out the moderation layer of its AT Protocol network under visible strain. An early moderation policy change alienated Black users who expected tougher action on bigotry even when it clashed with decentralized ideals, and the company later answered scale problems by quadrupling its contracted moderator corps. Opening federation to any operator raised the stakes further: enforcement decisions now touch independent servers, not just bsky.app.
The published proposals — user lists, reply controls, hashtag-related tooling — are the next step in that arc: shifting enforcement labor from Bluesky's payroll into community-run workflows. They land just as the network crosses 40M users, adds 'dislikes' to tune reply rankings, and has already rewritten its Community Guidelines to satisfy the EU's DSA, the UK's Online Safety Act, and the US Take It Down Act.
First-order effects
- Bluesky users and server operators get primitives — shared user lists, reply controls — that turn block/mute actions from individual settings into reusable community infrastructure, reducing how much enforcement bsky.app staff must perform directly.
- The package draws immediate criticism from commentators who note the labeling requirements would cover child-abuse images and DMCAed content while leaving Nazi content permissible in some jurisdictions, putting Bluesky's 'public commons' framing under dispute.
Second-order effects
- Combined with custom feeds and algorithms built on the same list primitives, the tooling advances a choose-your-own-moderation model in which third-party curators and federated server operators — not Bluesky employees — become the effective gatekeepers for swaths of the network.
- Rivals and observers watching Mastodon's ActivityPub ecosystem now have a direct comparison point for whether protocol-level moderation primitives can substitute for centrally staffed trust-and-safety teams.
Third-order effects
- If the model holds, accountability splits awkwardly: regulators under the DSA and Online Safety Act hold Bluesky legally responsible for systemic risk, while day-to-day enforcement is delegated to user-built tools whose quality no single party guarantees — a structural test of whether statutory duty-of-care regimes can coexist with distributed moderation.
The trend: Social platforms are shifting moderation from centrally staffed enforcement toward user-configurable tooling layered on open protocols, even as regulation keeps final accountability with the platform itself.