Moderation Policy

Version 1 — pending legal review before final publication.

What PUNCH removes

PUNCH removes content in the following categories. This list is exhaustive — content outside these categories will not be removed, restricted, or de-prioritized.

1. 1. Content that sexualizes, endangers, or exploits children.

   Zero tolerance. Reported to NCMEC where required by law. Accounts terminated immediately and permanently.

2. 2. Direct, credible, specific threats of physical violence against identifiable people.

   “I hate this politician” is opinion. “I am going to kill [name] at [location] on [date]” is a threat. We act on threats. We do not act on opinion.

3. 3. Doxxing.

   Posting another person's home address, phone number, workplace, family members, or other identifying information without their consent.

4. 4. Non-consensual intimate imagery.

   Sexual content involving identifiable people who did not consent to its sharing. Removed immediately on report.

5. 5. Content that is illegal in the country it is posted from.

   Including but not limited to: human trafficking, sale of weapons or controlled substances to private buyers, content that violates copyright after a valid DMCA notice.

6. 6. Spam and inauthentic behavior.

   Coordinated fake accounts, bot networks, large-scale automated posting.

7. 7. Direct incitement to imminent lawless action.

   Brandenburg standard — speech directed at producing imminent illegal violence and likely to do so.

What PUNCH does not remove

• Political opinions of any kind, including positions about wars, governments, religions, and movements
• Criticism of any company, politician, public figure, or institution
• Discussion of illegal activities (vs. coordination of them) — drug harm reduction, sex work safety, protest tactics
• Content that other platforms have throttled for “engagement” reasons, controversy, or platform-side political pressure
• Nudity in non-sexual contexts (art, breastfeeding, medical, body-positive) — we will not over-moderate the human body
• Strong language, sarcasm, satire, parody

How reports work

• Any user can report any post, account, or DM (DMs reported via report-with-context flow)
• Reports are reviewed by a human within 24 hours
• Both reporter and reported user receive a written outcome
• If your content is removed, you get the specific reason and the right to appeal
• Appeals are reviewed by a different moderator than the one who made the original decision
• Strikes are public to you. There is no shadow strike system.

What we do not have

• Shadow bans
• Reach throttling without notice
• Algorithmic demotion based on political content
• “Sensitive content” flags applied without your knowledge
• DM scanning. We cannot scan your DMs because they are Encrypted DMs. Our moderators have no access.

Transparency

We publish a quarterly transparency report listing: total content reports, total removals by category, total appeals, total reversals, total government requests, and total government requests complied with.