Product2 distinct publishers3 min readPublished
The account-level toggle keeps a user's posts out of Bluesky's Discover feed for people who do not follow them, but honoring it elsewhere on the protocol is voluntary, which leaves reach as the control on offer.
The Product Desk · Product desk
Compiled by The Product DeskSomething wrong?How this is made
The moment someone goes looking for this setting tends to arrive late. A reply written for followers has been lifted in front of an audience it was never meant for, and the author starts opening menus. Bluesky says the toggle can take up to an hour to fully take effect [5]. Paired with its all-or-nothing scope, that makes it a posture you set before trouble rather than a brake on a post already moving [15].
What the switch adjusts is ranking, not reachability. Posts on the network are public by default and Discover can currently surface any of them [7], so opting out narrows who gets served your posts in one feed while leaving them readable to anyone who opens your profile. That feed is worth something. Engadget calls Discover one of the main places people on Bluesky find accounts they do not already follow [12], which means the price of the setting is the post you did want to travel.
The most useful thing on the screen is the wording of the checkbox: "ask apps to hide my posts from algorithmic recommendations" [4]. The verb is doing real work. Guaranteed enforcement covers one ranker, Bluesky's own [14], and Engadget notes that posts can still turn up in other algorithmically curated feeds and services built on the AT Protocol [10]. Engadget also reads fully private accounts as unlikely to ever arrive on Bluesky [16], which leaves reach as the thing users are actually handed.
Teams tell themselves that users who worry about reach want fewer impressions. What users do is narrower: they post to the audience they believe they have, and they go quiet when strangers show up in the replies. Bluesky describes the change as one of a few updates meant to give people more control over their experience and help them find their people [11], which is a vaguer claim than "privacy setting" and a more accurate one.
Two axes are enough to grade any reach control a vendor ships you. Where enforcement lives: inside the code that does the ranking, or in the goodwill of whoever reads the flag. What granularity it costs: the individual post, or the whole account. This one sits app-enforced and account-wide, the cheap corner on both counts, which is close to the limit of what a public-by-default protocol can enforce today.
So the rollout answer for Monday has two halves. If your users need per-post control, the honest advice is two handles, one for the work that should travel and one for everything else. If they need a guarantee, the honest advice is that the guarantee stops where Bluesky's ranking code stops [14], and anything downstream is a courtesy. In a month, the number worth pulling is whether the people who switched it on are still posting as often as they were, which tells you more about whether the setting worked than any count of Discover impressions.
Ranked by verification strength, evidence, and original report placement.
Bluesky on Thursday introduced an algorithmic opt-out feature that allows users to stop their posts from appearing in the app's main Discover feed, and that prevents posts being shown to people who do not already follow them.
The tool is an account-level control, not a per-post filter, so enabling it prevents all of a user's posts from being shown in Discover to people who do not already follow them.
The opt-out is toggled on via a new option in the app's Privacy and Security settings.
The setting's checkbox is labelled "ask apps to hide my posts from algorithmic recommendations".
Bluesky says the change can take up to an hour to fully take effect.
The preference is recorded at the account level rather than only inside Bluesky's app, so it travels with the user to other apps powered by AT Proto; those apps have access to the information but still have to choose whether to respect it.
Distinct publishers with included, body-backed reporting in this cluster.
2 articles · August 28, 2026
2 articles · August 28, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
product
X builds the NFL a feed of its own to coax media buyers back1 distinct publisher
build
Bluesky names its infrastructure, and puts a token on the archive1 distinct publisher
build
A Bluesky starter pack is three records, and only the metadata edit is repeatable1 distinct publisher
science
3.2 million replies, 88 candidates, and the gender gap that prevalence metrics erase1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One source, relayed twice
Every checkable detail here — the settings location, the checkbox wording, the hour before it applies, the fact that the preference lives on the account — originates with Bluesky and reaches us through TechCrunch and Engadget. The feature itself is plainly real and in the app, so this is not thin reporting. What is missing is the single test that would settle the story: neither outlet opened another AT Proto client to see whether the flag is read there.
Live in one app, honored by nobody we can name
The toggle exists and any Bluesky user can flip it today, which is real adoption of a kind. Past that, the ledger is empty: no figure for how many accounts have used it, and not one named AT Proto app that reads the flag. The design deliberately hands that decision to third parties, and so far the story contains zero evidence any third party has made it.
The name promises a network, the feature delivers a feed
'Algorithmic opt-out' sounds like a switch that binds the protocol; what is guaranteed is one ranking system inside one app. The overstatement sits in the framing rather than in the journalism — TechCrunch spells out that other apps choose whether to comply, and Engadget goes further by reminding readers Bluesky has no private accounts at all. Small positive gap, and both publishers did most of the deflating themselves.
The subject wrote the only account
Bluesky is the sole origin of the facts and the beneficiary of the framing: a privacy-shaped headline that costs it nothing architecturally, since the network stays public by default and the obligation on other apps stays optional. The company also supplies the motive story — people who do not want to go viral — which both outlets adopt. On the publisher side, the pull is toward a quick, relatable consumer-feature story, visible in the accidental-main-character angle; neither has a stake beyond that.
Facts firm, consequences open
Nothing in the two reports contradicts the other, the feature is shipped, and the mechanics are specific enough to act on. The uncertainty is not about what happened but about what follows: whether any other client honors the flag, how much reach an account forfeits, and whether the private-data work TechCrunch mentions ever lands — where Engadget's 'unlikely ever' is a guess, not a finding.