Reviews and ratings on an adult directory: what actually protects the operator when an advertiser threatens to sue

When a provider threatens to sue over a review

Plenty of directories let clients rate a provider after a booking, and sooner or later one of those reviews provokes an angry response. Sometimes it is an email threatening a lawsuit. Sometimes it is a demand to take the review down, citing defamation. Rarely, it is an actual subpoena. Operators who have never faced any of these tend to assume they are all the same problem, and operators who have faced one tend to overcorrect for the next one that arrives.

They are not the same problem. An angry email creates no legal obligation at all. A formal takedown demand still creates none, on its own. A subpoena from a real court case does, and it is the only one of the three an operator is required to act on. Treating all three the same, by deleting on demand or by ignoring everything equally, gets the operator into trouble from opposite directions.

Delete every negative review the moment someone complains, and the rating system stops meaning anything. Providers who behave badly learn that a firm enough email makes criticism disappear, honest clients notice the pattern, and the feature meant to build trust in the directory starts working against it. Ignore a real legal document because the last ten complaints were empty threats, and the operator can end up personally on the hook for a case that was never really about them.

The rest of this piece works through what actually protects an operator by default, where that protection can crack depending on how the review feature itself is built, and what to do differently with a genuine legal threat versus an angry email that only sounds like one.

The default protection, and why it is not automatic forgiveness

In the United States, Section 230 of the Communications Decency Act is the reason a client who posts a false, damaging review is usually the one who can be sued for it, not the directory that hosted it. The law treats a platform as a distributor of someone else's speech rather than the author of it, so a court will not normally treat the operator as having published a claim it did not write, even when the claim turns out to be false.

That protection tracks a real distinction, not just a legal technicality: a review that says a booking left the client unhappy is an opinion, and opinions are protected speech regardless of how unfair they feel to the person reviewed. A review that asserts a specific, checkable fact, such as a claim about money taken or an appointment that never happened, is the kind of statement defamation law actually cares about, because it can be shown true or false. Moderation effort belongs on that second category, not on tone.

None of this depends on the operator being timid about it. Pulling a negative-but-plausible review the moment a provider complains loudly is not something the law requires, and doing it anyway teaches every advertiser on the platform that a firm enough message gets criticism erased, whether or not the review was true. A directory that folds on the first complaint every time ends up with a review section that only reflects who complains hardest, not who actually performs well.

The shield also has an edge worth knowing: it protects the operator from liability for what a user wrote, not for claims the operator's own staff adds on top, such as a badge or a note asserting something the platform cannot actually back up. If moderators add editorial commentary to a review rather than just deciding whether to publish it, that commentary is the operator's own speech and stands or falls on its own.

None of this needs a lawyer on staff to apply day to day. The team that already reviews listings before they go live can be handed a short written rule for reviews too: read for a specific, checkable factual claim, not for whether the review is unkind, and only pull content that fails that first test.

Where the shield can actually crack: how the review form is built

In 2008, a US federal appeals court ruled on a case that had nothing to do with reviews. Fair Housing Council of San Fernando Valley v. Roommates.com was about a roommate-matching website that required users to answer specific questions about protected characteristics before they could use the site. The court found that because the site itself designed the mandatory questions and pre-set answer options, it had helped create the resulting content rather than merely hosting whatever a user chose to type, and it lost Section 230 protection for that part of the site. An open-ended "additional comments" box on the same page kept its protection, because nothing about that field pushed users toward any particular content.

That case is about housing discrimination, not reviews, and it should not be read as a ruling on review features specifically. But legal commentators consistently point to the distinction it drew, between a platform that passively hosts what a user decided to write and one that actively prompts the content through its own design, as a caution worth applying anywhere a site collects user-generated text. It is a signal to watch for, not a settled rule that automatically applies to every structured form.

The practical read for a directory is straightforward: a review feature built around a star rating and an open text field sits on much steadier ground than one that adds dropdown menus or checkboxes prompting reviewers to select specific service details from a fixed list, because the fixed list, not the client, is doing part of the writing. The safer the platform's own contribution to the content, the more the review stays the user's speech rather than the operator's.

For an adult classifieds site specifically, there is a second reason to keep review categories neutral and operational, covering things like punctuality, communication, and whether the photos matched. Prompting or structuring reviews toward specific commercial-sex descriptions would not only raise the Section 230 concern above, it edges toward the facilitation exposure that a separate federal law creates for adult classifieds operators, because a form that generates that language is doing more than hosting a customer's own words.

Keep the structured part of the review to neutral, checkable operational facts, leave everything else to an open field, and moderate that open field for specific false factual claims rather than for how blunt the wording is. That single design choice does more for the platform's legal footing than any takedown policy written after the fact.

What a real threat looks like, and what an angry email is not

A cease-and-desist email, or a message threatening to "get lawyers involved," is not a court order, and an operator is under no legal obligation to remove a review just because someone asserts it is false. The operator is also not the right party to adjudicate a booking dispute between a client and a provider; the platform's job is to check whether the review makes a specific, verifiably false claim, not to referee who was actually right.

It is worth knowing that anti-SLAPP protection exists for exactly this situation, though only at the state level in the US, with no federal version. Many states let someone sued over a genuine, opinion-based negative review get the case dismissed early and recover their legal costs from the person who sued them, which makes a weak defamation suit over an honest review expensive for the party who files it. How strong that protection is varies a great deal from state to state, so it is something to be aware exists, not something an operator can promise a reviewer in advance.

On a plain demand letter, the useful response is short and consistent rather than improvised: acknowledge receipt, note that the platform reviewed the content, that it did not author it, and that it does not remove genuine opinions or unverified factual claims without an actual court order. Having that written down in advance, so a moderator does not have to compose a legal position from scratch under pressure, is worth more than any single clever reply.

That does not mean every negative review survives contact with a complaint. If a review makes a specific factual claim the operator can independently see is false, or that the reviewer cannot back up when asked directly, taking it down is the right call regardless of who complained, because continuing to publish something known to be false is a different problem from standing behind a genuine opinion.

Real subpoenas are rare next to the volume of angry emails a busy directory receives, and that gap is exactly why it is worth setting a high bar for reacting to a threat and a much lower one for reacting to an actual court document. The rest of this piece is about that second, less common case.

When a real subpoena arrives asking who wrote a review

To sue over an anonymous review, a claimant generally cannot get the author's identity from the operator directly. They have to file a lawsuit against an unnamed "Doe" defendant first, then subpoena the platform for whatever identifying account information exists, such as an email address or the IP address captured at signup. That is a different process from a law-enforcement subpoena tied to a criminal investigation, which follows its own separate rules.

Many courts do not let that kind of subpoena go through automatically. Before ordering a platform to unmask an anonymous user, a court will often require the person suing to show real evidence the underlying claim could actually hold up, not just that they are upset about the review, because identifying an anonymous reviewer by mistake cannot be undone once it happens.

The operator's job in that situation is not to fight the case on the reviewer's behalf, and it is also not to hand over identifying information the moment a threatening email arrives, since only an actual subpoena creates that obligation. The middle path most established platforms use is to notify the affected user that a subpoena has arrived and give them a real window, commonly something in the range of a week or two, to hire their own counsel and try to block it before the platform complies with what the document actually asks for, and nothing more than that.

This is one more reason the data-retention discipline worth building for identity checks is also worth building for reviews: keep only what the review feature actually needs, such as an account identifier, rather than a permanent archive of everything ever collected at signup. Write down, in one page, who reviews an incoming subpoena, how the affected user gets notified, and how long the window is before the platform responds. A policy that already exists before the first subpoena arrives reads as a business that thought this through in advance, not one improvising while a lawyer waits on the other end of an email.