False Claims on Facebook

The policy you are looking for is not in the index

A company that finds a false and damaging post about itself on Facebook nearly always starts in the same place: assembling proof that the post is untrue. Invoices. Delivery records. The order that was never placed. The customer who does not exist. Then the report goes in, and it closes with no action.

Nothing went wrong with the filing. The filing answered a question Meta never asks. Meta publishes one rulebook for Facebook, Instagram, Messenger and Threads — the Community Standards — and the index runs to roughly twenty-five policies. Coordinating Harm and Promoting Crime. Fraud, Scams, and Deceptive Practices. Violence and Incitement. Bullying and Harassment. Hateful Conduct. Privacy Violations. Account Integrity. Misinformation. Spam. Third-Party Intellectual Property Infringement. Locally Illegal Content. Not one of them is defamation. There is no category for “this statement is false,” because falsity is a question of fact that no reviewer can settle from a report form, and Meta does not attempt it.

So the work on Facebook is translation. You take what actually happened to the business and you find the conduct rule it breaks, if it breaks one. Some matters translate. A great many do not, and when they do not, the honest answer is available in the first hour rather than the third month.

“Our Community Standards apply to everyone, all around the world, and to all types of content, including AI-generated content.”

— Meta, Community Standards (index), transparency.meta.com, read 12 August 2026

That final clause earns its place in 2026. A fabricated screenshot of an invoice, a synthetic clip of a founder saying something he never said, a generated account of a transaction that never happened — all of it sits inside the same standards, and those standards still contain no rule against being wrong about a company. Meta stamps every policy with a change log, and the dates on this page are the dates those pages carried when I read them on 12 August 2026. Platform policy moves, sometimes quietly. Open the operator's current page before acting on anything written here.

The tier that decides most business cases, and the sentence inside it

Bullying and Harassment is where most companies file, because it is the only policy whose name resembles what is happening to them. It runs in four tiers. Tier 1 is universal and covers everyone. Tiers 2, 3 and 4 layer on further protection depending on who the target is — private adults, minors, limited scope public figures, private minors. An executive who is not a household name sits in Tier 3.

Tier 3 is also where Meta wrote the exception that ends most business complaints:

“Negative character or ability claims, except in the context of criminal allegations and business reviews against adults.”

— Meta, Bullying and Harassment, Tier 3, change log dated 22 January 2025, read 12 August 2026

Read the exception rather than the rule. Private adults are protected from negative claims about their character or their ability — except where those claims arrive as criminal allegations or as business reviews. Meta has written business reviews out of harassment protection, by name, on its own page. A post saying the owner is a crook who takes deposits and never delivers is a negative character claim in the setting of a business review. It is excluded in terms.

The same tier carries a second exclusion built the same way: “Expressions of contempt, disgust, or content rejecting the existence of an individual, except in the context of criminal allegations against adults.” Between them, those two lines dispose of the overwhelming majority of what a company arrives wanting to report. Reporting a negative business post as bullying will fail, and it will fail without ever telling you why. That is not a neutral outcome either: it spends the days you had for preserving evidence, and it teaches the person posting that reporting is the move you make.

The Tier 1 lines that reach a real case

Tier 1 applies to everyone regardless of status, and its lines are about behavior rather than assertion. Three of them turn up in genuine business matters, and each of them is indifferent to whether anything said about the company was accurate.

• Repeated unwanted contact. Contact that is repeated, contact that is sexually harassing, and contact directed at a large number of people with no prior solicitation.
• Threats to publish private details. Quoted in full below, because it is the line a campaign against a business crosses most often.
• Calls for, or statements of intent to engage in, bullying and/or harassment. The post that says everyone go and tell them what you think is a different object from the post that merely says the company is bad. The first one is addressed by the policy. The second one is not.

“Threats to release an individual's private phone number, residential address, email address or medical records (as defined in the Privacy Violations policy).”

— Meta, Bullying and Harassment, Tier 1, change log dated 22 January 2025, read 12 August 2026

An attack on a company frequently escalates from criticism to something aimed at a person: where the founder lives, the phone number of the manager who handled the complaint, the threat to publish either. The hour that happens, a matter that had no route acquires one. Meta also removes directed mass harassment that targets an individual through personal surfaces such as an inbox or a profile, and states that it will disable accounts engaged in mass harassment as part of adversarial networks. None of that turns on truth. All of it turns on conduct, which is the reframe that makes the difference on this platform.

Who has to press the report button

Several of these protections are enforced only when the person targeted confirms that the content is unwanted. Meta's own wording is that it requires self-reporting in certain instances “because it helps us understand that the person targeted feels bullied or harassed,” and where somebody other than the target may act, the policy names “the target or an authorized representative of the target.”

That phrase does real work. Counsel or a designated representative can file. An unrelated well-wisher cannot — not the customer who spotted it, not the supplier who forwarded it, not the loyal employee with a personal account. Where the target is an individual rather than the company, the report that carries weight is the one the individual confirms. A great many reports die at this step and the reporter never learns that is why, which is the single most invisible failure mode on Meta's platforms.

Reporting is also a notified act. Enforcement generates a notification, the poster learns something was removed, and someone determined to keep attacking will edit, repost or open something new. Anyone who intends to hold on to the record should capture it before the first report rather than after: full-page captures carrying the URL, the handle and the date, the thread around the post, and the account's other posting. Once content is removed, it is removed from your evidence too, and Meta is not a records service.

“Public figure” means one thing to Meta and another in a courtroom

Meta defines the term on its own page: “state and national level government officials, political candidates for those offices, people with over one million fans or followers on social media and people who receive substantial news coverage.” A separate “limited scope public figures” category covers people whose fame is confined to their activism or journalism, or who became famous involuntarily.

Hold that definition against the executives of ordinary companies and almost every one of them is a private adult to Meta. The chief executive of a substantial regional business, with a few thousand followers and trade-press coverage, is not a public figure under this rule. That sounds like good news for a moment, and then it is not, because being a private adult puts the matter in Tier 3 — the tier carrying the business-review carve-out quoted above.

Whether that same person is a public figure for the purposes of a defamation claim is a different question, decided by a court on a legal test, and it is a question for a defamation attorney rather than for me. What matters on this page is narrower and entirely practical: a business owner who has been told “he's a public figure, nothing can be done” has usually been handed a platform rule and a legal test welded together into one sentence, and whoever welded them checked neither. They are separate systems. They ask different questions and they can answer differently.

Remove, reduce, inform — and only the first one is removal

In a policy report filed with the New York Attorney General for Q3 2025, Meta describes its enforcement in three parts: it removes content that goes against its policies, it may reduce the distribution of problematic or low quality content that does not violate them, and it informs people with additional context.

Those three have to be kept apart, because plenty of vendors will not keep them apart for you. Reduced distribution means the post is still published, still at its address, still visible to anyone who holds the link or lands on the profile, and still available to be screenshotted into a group chat. A dashboard showing that reach fell is not a takedown, and a report that a post was “actioned” can mean any of the three.

The surfaces matter as well. Meta states that the protections of Tiers 1 through 4 are also enforced on pages, groups, events and messages, so a Page built specifically to attack a company is assessed under the same lines as a stray comment — there is no separate and easier rule for a hate Page. And a complaint about someone's home address or medical details belongs under Privacy Violations rather than Bullying and Harassment. Different policy, different test, and picking the wrong one is how a viable report is closed as unfounded.

The order I work in, and the four things that will not work

Sequence matters more than effort here, because two of these steps are irreversible once taken.

1. Preserve first. Before any report, before any message to the poster, before anything that tells him he has been noticed.
2. Find the line, then write to the line. Not to the injury. The reviewer is matching content against a policy sentence, so the report has to name the sentence and show the match.
3. Ask whether this is one post or a pattern. A pattern is a different and stronger case, and it is made out of accounts rather than content.
4. If nothing fits, say so early. Then the remaining work is the search result that carries the post, the response, and the record — which is a smaller answer than the reader wanted and a truer one than the vendor gave.

Four things that will not work on Facebook, stated plainly so nobody spends a month discovering them:

• Reporting a negative business post as bullying. Carved out by name in Tier 3.
• Arriving with proof of falsity. No policy line takes falsity as its test, so the strongest evidence file in the world has nowhere to be filed.
• Filing on behalf of a named individual from the company account where the policy requires confirmation from the target or an authorized representative.
• Sending Meta a demand letter and waiting. Notice to a platform that content is false does not create an obligation to act on it — that is Zeran v. America Online (4th Cir. 1997), and it means “I told them and they left it up” describes a lawful outcome rather than a wrong.

Frequently Asked Questions

Published 2026-08-12