One Fandom Block List Can Quietly Decide Who Gets Heard

The artifact to keep in view is a CSV file. It is plain text arranged in rows, built so software can read a list of accounts without needing the story behind each one. In 2015, Twitter formally let users export and import block lists in this format. One person could do the miserable work of identifying hostile accounts; everyone else could import the result.

This sounded humane because the alternative was absurd. A user facing coordinated harassment had to inspect accounts individually, open menus individually and block individually, while the people targeting them could coordinate elsewhere and arrive together. Shared blocking turned defense into a bulk action.

It also turned the CSV into a tiny court record with the case files removed.

The block list became infrastructure

GGAutoblocker, created by developer Randi Harper during GamerGate, showed both the need and the problem early. GamerGate was a networked harassment campaign wrapped in an argument about games journalism, and its targets faced waves of abuse that platform-level enforcement did not reliably stop. GGAutoblocker used social-graph signals, meaning patterns of who followed whom, to identify accounts connected to prominent GamerGate hubs and make them blockable in bulk.

The tool did not need to prove that every included account had personally sent harassment. That was the point. Waiting for each account to cross an individual evidentiary line meant waiting for the harassment to arrive, then asking its target to document it for a company whose reporting interface already treated abuse as a sequence of isolated posts. The blocker worked upstream.

Association became a risk signal.

Public reporting on GGAutoblocker also documented the predictable objection: social proximity is messy. People follow hostile accounts to monitor them, argue with them, report on them or keep track of a community in which they have become a target. An account can sit near bad actors in a follower graph without behaving like them.

The CSV cannot hold that distinction. Its clean rows are useful because they contain almost nothing.

Shared lists have since become ordinary across fandom spaces, especially where users expect dogpiles, ship wars, racist harassment, anti-trans campaigns or sustained attacks on fan artists and writers. On Bluesky, moderation lists let a creator collect accounts that other users can mute or block together. Third-party tools around social platforms offer similar functions, while spreadsheets and screenshots still circulate when official tools are missing.

The platform supplies the button. Fandom supplies the policy.

A username is made to carry the whole case

Most fandom block lists move through trust networks. A well-known fan, moderator, artist or community organizer posts a list. Mutuals repost it. People import or subscribe because they trust the maintainer, because trusted friends already did, or because reviewing hundreds of accounts would defeat the purpose of bulk blocking.

That last point matters. A shared list promises to save attention, so any system that expects every subscriber to audit every entry is pretending the feature does not exist. Delegation is the product.

The maintainer may collect screenshots, direct messages and reports from other users. They may also add accounts after seeing a hostile reply, a profile slogan, a follow, a repost or inclusion on another list. Standards vary because there is no common standard. One list may target documented harassers.

Another may include anyone interacting with a disliked creator. A third may collapse political disagreement, personal conflict and credible safety concerns into the same column.

Then the evidence falls away as the list travels. A screenshot thread becomes a username. A username becomes a row. The row is imported by people who never saw the original conduct, and later copies may not retain even the maintainer’s short explanation.

GGAutoblocker made its inference legible enough to argue about: it was using network association as a proxy for risk. Many fandom lists are less explicit. Subscribers see a title such as an anti-harassment list or a warning that listed accounts are unsafe, but not whether each person threatened someone, followed the wrong account, liked a post during a pile-on or got added after a private dispute that outsiders cannot evaluate.

The compression is not an accident or merely poor recordkeeping. It is how the list scales.

Volunteer moderators inherit state power without state resources

A shared block list does not ban anyone from the platform. That distinction is real. Blocking changes who can reach a participating user, while platform moderation can remove posts, restrict distribution or suspend an account.

Inside a tightly networked fandom, though, coordinated blocking can shape practical access to the community. If enough prominent accounts subscribe to one list, a listed user may lose replies, mutual connections and routes into conversations. Recommendation systems also learn from interaction patterns, though platforms disclose too little to calculate how any particular wave of blocks affects distribution. The result can feel like expulsion even when no central administrator issued one.

The person maintaining the list now controls a moderation layer. They decide what counts, whose reports carry weight, whether old entries expire and what happens when someone disputes inclusion. The job can consume hours of unpaid labor, expose the maintainer to retaliation and force them to inspect the same harassment everyone else is trying not to see.

Platforms benefit from this arrangement. Users build the lists, assess danger and absorb appeals; the company keeps the engagement and can still point to user-controlled safety tools. Twitter’s importable CSV and Bluesky’s moderation lists did not invent community defense. They packaged it as a feature while leaving the hardest decisions to volunteers.

There is little direct money in maintaining a fandom block list. The economic value sits elsewhere. Better self-defense can keep targeted users on a platform, and volunteer curation reduces pressure on paid trust-and-safety teams. Third-party moderation services may charge for broader controls, but the raw material often remains unpaid reports from people documenting their own harassment.

The spreadsheet is free. The work inside it is not.

The list can become the harassment tool

The same mechanism that protects a target can identify one. Public moderation lists may reveal clusters of trans users, activists or members of a specific fandom to anyone who finds the list. A hostile curator can label those people as dangerous, collect them in one place and hand followers a ready-made directory.

This is the structural weakness of user-generated moderation. The platform can let people choose whom to trust, but it cannot assume that every curator acts in good faith. A list title is metadata, not proof. Popularity is not an audit.

Context also decays. Accounts change hands. People change behavior. Disputes resolve or become impossible to reconstruct after posts disappear.

A copied CSV freezes an earlier judgment, while a subscribed list gives the maintainer continuing power to add accounts later. Those are different products wearing the same safety language: one is a snapshot, the other is a live rule imposed across many accounts.

GGAutoblocker’s controversial edge was visible because its selection logic was discussed in public reporting. A contemporary fandom list can conceal more power behind less machinery. The curator may be anonymous, the evidence thread may vanish and subscribers may never notice when the list’s scope shifts from preventing harassment to enforcing factional loyalty.

None of this makes bulk blocking illegitimate. A person under attack does not owe each incoming account a hearing before closing the door. The problem begins when a private safety choice becomes portable governance for thousands of other people, while still being treated as too informal to require records, limits or accountability.

Better moderation costs more than a button

A less reckless shared-list system would preserve reasons for inclusion, distinguish documented conduct from network-based risk and show when an entry was added or reviewed. It would let subscribers choose categories rather than accept one undifferentiated verdict. Appeals could go to more than one maintainer, especially when the list functions as community infrastructure rather than one person’s private boundary.

Those controls cost time. Evidence needs storage and protection. Appeals expose moderators to more contact. Review dates turn a one-time list into ongoing maintenance.

Multiple curators reduce unilateral power but create arguments over standards, access and responsibility. Platforms routinely avoid these costs by offering a bulk button and calling the result user choice.

The right comparison is not between shared lists and some frictionless, perfectly fair system. It is between forms of labor. Without a list, targeted users repeatedly block harassment themselves. With one, a curator performs that work for everyone and gains authority as a result.

With platform-run moderation, the company must hire people, write enforceable rules and accept responsibility for mistakes.

The CSV from 2015 still explains the bargain. It moves quickly because it is small. It stays small because the testimony, uncertainty and possibility of change have been cut out.

Questions people ask

What is a fandom block list?

A fandom block list is a shared collection of social-media accounts that users can block or mute in bulk. Lists may target documented harassers, spam accounts or participants in coordinated attacks, but the inclusion standard depends on the maintainer and may not be visible to subscribers.

Who maintains shared block lists?

They are usually maintained by volunteer fans, community moderators, artists, activists or anti-harassment organizers. Platforms provide import, subscription or bulk-blocking tools, but maintainers often perform the investigation, updates and dispute handling without pay or institutional support.

Can someone be added without harassing anyone?

Yes. Some lists use association signals such as follows, reposts or participation in a hostile network, while others rely on reports that subscribers cannot inspect. That can stop an attack before it reaches someone, but it can also catch researchers, observers, critics or people added through a personal dispute.

Are shared block lists the same as platform bans?

No. A block list changes interactions for the users who adopt it; a platform ban restricts an account across the service. In a close fandom network, however, widespread adoption can still cut a listed person off from much of the community without a formal moderation decision or appeal process.