On a few occasions I have had well thought out legit reviews initially rejected with the instructions to describe more specific details of our encounter. I am fairly vanilla and consistent on my preferences and performances. The last time this happened I waited 15 minutes and resubmitted my review and within 15 minutes it was approved. Maybe AI was involved here as well.
Recently, I’ve read reviews which appear as having been written by AI. I use it often, not for reviews, so I can discern the style. I wish I kept the links to share and get views of others. What concerns me is that AI makes easy to write fake reviews quickly. In two cases I noticed that there were two reviewers who posted reviews about the same two providers over a short time period. In any event, I wonder if you noticed an increase in fake reviews on TER.
Some years ago I saw a review here of a girl I'd seen and reviewed on another site. When I read it, it was clear the person had cut and pasted MY review and presented it here as his own. I informed TER and the review was taken down and the handle was banned. Who knows how often that still happens today and it's probably easier than using AI.
There have always been fake reviews that I have noticed. I haven’t noticed an increase. In all likelihood, those willing to use AI are the same ones willing to previously write fake reviews without AI.
I do when I see my fake reviews. TER gives out free VIP days to write reviews. Fake reviews have always been an issue. I think there should be a verification process to ensure reviews are accurate. 🤷🏽♀️ I utilize all TER’s tools. Keeping up to date by reading your reviews is always important.
I’ve used AI after I’ve written the review to parse grammar and flow. Wish I had it in college 😂
I was just going to post a thread about this
I use ChatGPT a lot, and it has a certain way of writing, and many reviews for the NYC providers are definitely written by AI
The reviews make it seem like a perfect storybook date when most men do not know even know how to write, let alone use flowery descriptions for an appointment
Also seen a huge bump of reviews from reviewers who have only seen that one provider or her and only her friends,
And hate to say it, because these ladies are my favorite, a lot of the reviewers and reviews are for Russian women
Plus I was going to book a woman who is active and twitter and has a website but her new reviews are from one or two time reviewers, and she asked for 30% deposit through bitcoin
Definitely there needs to checked
Recently, I’ve read reviews which appear as having been written by AI. I use it often, not for reviews, so I can discern the style. I wish I kept the links to share and get views of others. What concerns me is that AI makes easy to write fake reviews quickly. In two cases I noticed that there were two reviewers who posted reviews about the same two providers over a short time period. In any event, I wonder if you noticed an increase in fake reviews on TER.
I asked about this a month ago about provider I was considering seeing. Reviews seemed to mix up , "they and them " . It was hard to tell who the review was written by.
It's the other way around - an Ai review is far less likely to contain grammatical errors and might sound more formal and lack specific jargon.
Remember, Ai llms are just next token predictors. They don't have state and to emulate state they feed the entire previous generation back into it. Ie, the longer a piece of text - or a review - is, the more factual discrepancies you are likely to see. Something like you arrived at 5pm but then down the road few paragraphs later it will say the clock was showing noon. The bigger the context window size, the harder it is to track all of it without actual state
Ai generated reviews I'd think are either extremely generic or like the most penthous-y reviews ever, with plot twists and turns, drawing content form fiction literature training models.
Now of course nothing prevents someone to generate an Ai skeleton and then edit it, fixing mistakes and making it sound like an actual review by a normal person. However, unironically, people will only get lazier using AI generation, so I doubt many would do this.
-- Modified on 7/11/2025 6:39:37 PM
-- Modified on 7/11/2025 6:51:54 PM
What if the person using it, thinking something like ChatGPT, instructs it to make a few errors?
Yes, it's possible. But you'd have to be very specific as to what errors it has to make and then have them consistently replicated across all "reviews".
That's another thing I neglected to mention. Consistency. Us humans have developed distinct writing styles, with certain grammar, diction, phrasing, injecting references, and so forth. Along with any grammatical issues.
In order for Ai to have this consistency, you need to constantly feed previously generated reviews to it. Maintaining this consistency imo isn't trivial and requires some work that imo most people who would create such reviews, are probably not wiling to put into. After all, if you spend 30 minutes cleaning up inconsistencies and writing/editing better prompts, you could've written a fake review yourself in that time.
Even before llm Ai blowing up, id see accounts with reviews were being written in completely different styles, with completely different grammar, punctuation and phrasing by thesame people. And I'd immediately stop trusting those accounts. People just don't switch their writing styles 180 degrees like that. Usually it's an indication they just copied a generic review form elsewhere and just substituted correct names in.
I doubt most people would go to the lengths of trying to create consistency. But you can probably pick a few words, like improper use of your, and have that as a repeated mistake in a review. Someone willing to put in the detective work likely will figure it out. But I’m not convinced most clients put in the necessary work to figure these things out. In fact, I think the average person is somewhat lazy about it and that’s why these things work. The average TER user isn’t even posting on the boards. Look how few of us there are compared to actual accounts. I would say most of us on the boards are doing more than the average user. I could be wrong. Just my hunch.
Can't say I disagree with anything you wrote.
Yes, it's possible to do what you said but it's far less likely for someone to do so precisely because of what you said. Average user doesn't read into reviews, average user doesn't check users histories or do any kind of detective work. I do want them to read real reviews though and get educated on that matter, via any means necessary.
My only point was that a review containing things like pronoun errors and/or verb tense errors, is less likely imo to be Ai generated.
Why would a hobbyist write a fake review? It will not get the Escort to quickly jump his bones!!!! Any Hobbyists that does that to me.......will make me wanna question him a million times.
It's pretty obvious why. To get free VIP days.
Nothing is obvious to Kitty.
On a few occasions I have had well thought out legit reviews initially rejected with the instructions to describe more specific details of our encounter. I am fairly vanilla and consistent on my preferences and performances. The last time this happened I waited 15 minutes and resubmitted my review and within 15 minutes it was approved. Maybe AI was involved here as well.
I have spoken to several gentleman who have said they use AI to write reviews now because of the amount of “juicy” details required for posting. Some don’t know how to write that sort of thing. I have also seen a few providers who are ranked very high in my area pre write reviews with AI, then have the client post the review before the session starts, and if they do not then the provider terminates the session. This is why the top 100 is not really accurate anymore. Forcing someone to write a review while you watch them or pre writing an AI rev for them to post seems unbelievably desperate.
Did you "see" this behavior because you were doing a duo with those high ranked providers in your area? If not you didn't see anything.
As for the customers, if they are getting that done to them why are they not writing or posting about it? Once the appointment is terminated they have nothing to lose by reporting it. Moreover, they could simply allow the provider to force them to submit it during the session and then report that to TER and have the review removed and the provider banned. I'm guessing that would be worse for the provider (who perhaps thinks their ranking on TER is critical for their business) than not getting a review or having a some reviews that don't give some 10/10 that most of us don't put much weight on or believe if most of the reviews are that.
I think the message is getting lost playing telephone with Jemma, likely the providers offer a discount for the review and make the client post it when they first arrive. And then if they refuse in person, then they get kicked out. So the reviewers would have no incentive to report the behavior to TER because they would be outted/banned and no more discounted sessions with that provider.
Yes, if the clients are complicit in the fraud/rule violation then we get what we get. But with as many providers to see as there is, the large majority of which don't play that type of silly game from my experience, there really is not any justification for playing along.
Yes, if you're providing real world information and have a bit problem if it gets out then that is a bit on you. If that makes you the slave to the extortionist, well that part is on you as well. But outing a client is a very bad move on the part of a "high ranking" provider so I suspect the risk is low. Getting told to leave and not come back more likely but so what. Book someone else. A provider that needs to extort good reviews likely is not that good.
But simply making unspecified claims about something happening doesn't help anything. Jemma claims to have seen this happen. WHO did she see doing this? That would be helpful. That AI was or was not involved here is completely beside the point. Nor do I buy that the juicy details are so difficult to cover that someone needs an LLM to write a review TER can approve.
I have had several shy clients admit to using AI for the details part. Some are uncomfortable writing things like that for some reason. The AI reviews I was referring to have seen, and yes I did see it, these were pre written using AI by the provider, then sent to the client to copy and paste. As far as the specific providers, outing them would do nothing but cause drama, so instead I removed myself from their work group and no longer associate.
I did “see” it yes, personally, and I no longer work with that individual in the area. I also went to a provider mixer and witnessed it. Bottom line is when reading them take into consideration it may not all be accurate or even written by a person. They are there to verify a provider is legit, the juicy details will change from person to person