Shadowbanner Moderation Policy

Last Updated: 14 August 2025 Subscribe to us on Bluesky

Important Disclaimer

This section explains our relationship to Bluesky’s official moderation system and the limits of our authority.

• We are not affiliated with Bluesky Social, PBC.
• We do not replace Bluesky’s official moderation team or tools.
• Posts that violate Bluesky’s Community Guidelines — including threats, extremist content, and spam — should be reported directly to Bluesky via the in-app reporting system.
• If we receive a report that we believe violates Bluesky’s Community Guidelines, we may also submit a report directly to Bluesky’s moderation team using the same in-app reporting tools available to all users. This would be in addition to applying our own labels.
• We cannot deactivate or suspend accounts, or take down posts. Our capabilities are limited to applying labels, hiding labeled content, or blocking accounts/posts for subscribers to our labeler only. Posts remain viewable to anyone not subscribed to our labeler.

Purpose

Shadowbanner is an independent Bluesky moderation service focused on documenting and labeling harmful narratives, scapegoating, and bias-driven attacks — especially those that undermine marginalized communities, distort historical or ongoing atrocities, or excuse state and interpersonal violence.

We moderate using two primary tools:

• Labels applied to posts or accounts that meet our documented criteria.
• Moderation lists that group accounts meeting specific, objective inclusion criteria.

Our approach prioritizes clarity, professionalism, and factual grounding, ensuring that labels and lists serve as an accurate public record rather than a tool for personal attacks. The goal is not to silence disagreement, but to identify rhetoric that misleads, dehumanizes, or deflects accountability, and to make that information clear and accessible to users.

Labels are applied transparently, proportionately, and consistently so users can make informed choices about what they see and engage with. We start with post-level labels before considering account-level actions, and we operate within Bluesky’s Section 7 Labeler Guidelines.

Scope of Labeling

We apply labels to content that:

• Scapegoats marginalized groups.
• Spreads harmful disinformation targeting vulnerable communities.
• Excuses or justifies state or interpersonal violence.
• Mocks, dehumanizes, or delegitimizes victims of violence.
• Uses bad-faith tactics to derail discussion.
• Engages in coordinated spam or trolling that undermines healthy conversation.

We start with post-level labeling and only escalate to account-level labels when there is a clear and consistent harmful pattern.

Use of Moderation Lists

In addition to applying labels, Shadowbanner also maintains moderation lists. These lists may be compiled manually or generated automatically based on objective criteria, such as:

• Following specific public accounts
• Including certain keywords or phrases in a bio
• Meeting a combination of verifiable interaction patterns and follow activity

Some lists are automated (marked with 🤖) and are automatically updated on a daily schedule without manual review of each account.

Current Moderation Lists

1. Destiny Orbiters 🤖 Criteria: Accounts that follow Destiny (Kick streamer). Purpose: Allows subscribers to mute or block accounts that follow Destiny, or avoid seeing their content. Automated and updated daily. Context: Intended to help surface accounts that may engage in contrarian debates, reply frequently to pro-Palestinian/anti-genocide posts, or target journalists who write about Destiny.

2. Singal Subscribers 🤖 Criteria: Accounts that follow Jesse Singal. Purpose: Allows subscribers to mute or block accounts that follow Jesse Singal, or avoid seeing their content. Automated and updated daily. Context: Intended to help surface accounts that engage with or subscribe to Jesse Singal’s work.

3. KHive Accounts 🤖 Criteria: Accounts that follow Kamala Harris and have KHive-related keywords/hashtags in their bio. Purpose: Allows subscribers to mute or block KHive-affiliated accounts, or avoid seeing their content. Automated and updated daily. Context: The KHive is an online political community of supporters of Kamala Harris. This list is intended to help surface accounts that actively identify with or participate in KHive-related online activity.

4. Gaza Disinfo Follower Criteria: Accounts that follow one or more profiles documented for sharing Gaza-related disinformation. Purpose: Allows subscribers to mute or block accounts that follow such profiles, or avoid seeing their content. Context: These followed accounts have been documented as promoting narratives that attempt to delegitimize Palestinian suffering and dismiss evidence of war crimes — such as the “Gazawood” narrative used to discredit genuine documentation from Gaza.

Global List Disclaimer: Inclusion on any Shadowbanner moderation list is based solely on the objective criteria for that list. It does not imply that every account on the list engages in the behaviors mentioned in its context section. Moderation lists are tools for allowing subscribers to mute or block accounts, or avoid seeing their content. Presence on a list is not a statement of moral judgment.

Alignment with Bluesky Section 7 Guidelines

We follow the Bluesky Community Guidelines – Section 7: Labeler Guidelines. This means our labels are:

• Clear and non-abusive — targeting the content, not the person.
• Consistent and evidence-based — applied using our documented label definitions.
• Proportionate — starting with posts, escalating only when necessary.
• Transparent — we make our label definitions public, and we continually review them for adequacy, adding new ones when needed. If your post or account has been labeled, your first step should be to submit an appeal in the Bluesky app (see the Appeals section below for instructions). If you aren’t satisfied with the outcome of the in-app appeal, you can then contact us via Bluesky DM or email at [email protected] for further discussion.
• Impartial — applied regardless of political affiliation when criteria are met.

Limitations

We generally do not label:

• Content that is only annoying, partisan, or “cringe” without fitting harmful narrative criteria.
• Outdated posts unless the same harmful narrative is still actively promoted by the same user.
• Reports without specific posts (except in cases of a clear ongoing pattern).
• Satire, parody, or quoted material presented for criticism — unless it spreads the harmful narrative without challenge.

Handling Edge Cases

How we approach posts where the intent, tone, or context is less clear.

• Intent vs. Impact: We label based on harm caused, not just intent.
• Satire & Parody: Exempt unless indistinguishable from sincere harmful content.
• Historical Content: Labeled only if representative of the account’s ongoing behavior.

Reporter Guidance

Tips for making reports that help us review them efficiently.

• Focus on specific posts rather than accounts unless the harm is consistent.
• Prioritize recent content — older posts may not reflect current views.
• Include a short comment explaining why the post meets our criteria.
• Check our label list to see if the behavior matches a defined category.

Appeals Process

We’re human and sometimes make mistakes — the appeals process exists to correct them. The steps below explain how to request a review if you believe a label was applied in error.

• First and fastest option: Submit an appeal in the Bluesky app using the built-in appeal system:
• Go to your profile.
• Tap the “labels on this account” link.
• Tap “Appeal” next to the label.
• Add your explanation.
• Press Submit — this sends the appeal directly to us.
• If you aren’t satisfied with the outcome of the in-app appeal, you may:
• Send us a Bluesky DM.
• Email [email protected] with the label, post/account link, and explanation.
• Appeals are reviewed in the order they’re received. We aim to respond within 2 business days.
• Harassment of our moderators or targeting them for their moderation decisions will not be tolerated. Such behavior will be reported to Bluesky’s moderation team and may result in the account being blocked by our labeler.

Privacy & Data Handling

What information we can and cannot access about reporters and labeled accounts.

• All report and label data is stored on Bluesky’s PDS (Personal Data Server).
• We do not have access to any information about an account or reporter that isn’t already available through the public AT Protocol.
• We do not have access to personal details such as your real name, email address, or age.
• We do not publicly share reporter information.
• We do not collect, track, or store private data unrelated to the content reported.

Labeling Without Bias

Labels are applied impartially. Personal disagreements, political leanings, or affiliations do not influence labeling decisions.

Moderation Workflow

How reports are processed and the tools we currently use.

• All labelling decisions are made manually and not reviewed by AI.
• We currently use scripts to maintain certain moderation lists but do not use scripts to apply labels.
• We may explore automation in the future to streamline review processes without replacing manual oversight.

Policy Updates

When and how we notify users about changes to these guidelines.

We may update these Guidelines periodically. We will always post the revised version here and we will notify you through a post from @shadowbanner.bsky.social to communicate any material changes.

Thank you to everyone who takes the time to read, understand, and work within these guidelines. Your reports, feedback, and support make Shadowbanner possible.