AdultFriendFinder, reviewed on the only question that matters: the contract already answers it
Buyers want to know whether the people replying are people. The terms of use answer that, on the open web and without a login — the operator runs AI chat itself, and asks you to waive the claim for mistaking one for a person.
Nobody searches for this platform wanting a feature list. They want one thing settled before a card number goes in: whether the people replying are people. The reflex here is to report that the operator says nothing and move on. That reflex is wrong. The answer is published, in the contract you accept at registration, and it has been on the open web all along. For adults, 18+ — the only thing examined closely below is a company’s paperwork.
So the review runs on one axis: what does this operator commit to, in writing, about who is on the other end?
What the contract says
The document most reviews never open
A single unauthenticated request returns the full terms of use — roughly 73,000 characters of readable text, no account, no cookie wall. Every quotation below comes from it. The standard complaint about this market is opacity; on the contract, it does not hold. The problem is what it contains.
The platform runs the chatbots
The terms carry a dedicated section on AI features, covering AI-generated content, interactive story systems and AI chats. It calls AI-generated content, including messages from AI chat, “a fictional simulation powered by artificial intelligence”, and users agree not to rely on any output as “a real-time communication from a specific human being”. A separate beta section lists among the services offered “AI Virtual Friends, AI Friends, Matching systems, AI image generation and AI chatbots”.
Read that against the query that brings people here. The answer is not silence, and not a denial: some replies are generated, sold as entertainment, with the risk of confusing one for a person placed on the reader.
Two limits, stated honestly. The terms do not say member profiles are fake, and neither does this review — nobody outside the company can measure it. The document also never says how many users meet AI features, or whether they are on by default.
What the contract declines to promise
Three clauses close the exits. First, screening: the terms state in capitals that the company has “no obligation to screen our users or members”, inquire into their backgrounds, or verify their statements, and separately that it runs no criminal background screenings. That is a published verification position — an express refusal of one. The only mechanism offered is a report abuse page, used after the fact.
Second, moderation. The company does not guarantee prior review of AI-generated content, reserving a right without an obligation to remove outputs. Nothing commits it to labelling an AI message as one.
Third, the waiver. Users accept that outputs “may appear realistic or emotionally resonant” while being generated “without human verification”, and waive claims for harm from “misunderstanding, misidentification, or emotional reliance”. A contract that anticipates misidentification by name has told you what it expects to happen.
The money clause, also published
Reviewers who send readers inside the paywall for the renewal terms are pointing past an open door. Subscriptions renew automatically on expiry, for a period equivalent to the initial term, at the price then in effect, with auto-renewal switched off from the billing history page. Twelve US states get a three-business-day cancellation right. The price is the one fact withheld.
The security record
Have I Been Pwned holds a verified breach dated May 2015 covering 3,867,997 accounts, classed as sensitive: dates of birth, IP addresses, locations, relationship statuses, sexual orientations. An eleven-year-old incident, cited as history rather than current assessment — but on a platform of this kind, the fields are the point.
Against the category
Feeld and Pure publish verification figures that are self-serving and unaudited, picked over in our comparison of their verification and consent controls — but they name a product, and a reader can argue with a number. This contract names AI chatbots as a service and no verification product at all. Not two points on one scale.
Pros and cons
Verdict
Our assessment: approach this one knowing what you have been told. Read the AI section and the renewal clause before you subscribe, keep the first subscription monthly, and pay with a method that has a dispute path. Treat a refusal of a brief live video call as the answer it is, and meet somewhere public the first time.
The wider point cuts against this desk’s own habits. An operator that publishes nothing has declined to sit the test. This one sat it and gave the result in its lawyers’ language — more useful, much less comforting, and found only by opening the document.