Dating in times of AI

You say artificial intelligence is taking over dating platforms. How and for what purposes is AI used?
Dania Schiftan, sexologist and psychotherapist: The possible ways to use AI on dating platforms are manifold. You can use it to find good matches faster by searching the dating platforms for certain criteria. With apps like PhotoAi.Me, for example, you can optimise your profile picture. With programs such as ChatGPT, you can answer the questions about yourself in your profile in a particularly original way or they can suggest introductory questions that you can use to break the ice when you chat. In these cases, the AI is sort of acting as a flirting coach. Apps such as Mila Dating or CupidBot are primarily aimed at men and are designed to help them communicate with their matches by imitating the style of the person they are looking for.

Is it even possible to build meaningful relationships this way? It kind of sounds like you’re cheating on the other person.
If I use ChatGPT, for example, as a sparring partner to become a more interesting conversation partner, that’s not necessarily cheating. Otherwise, a course on the art of speaking or reading a flirting guide would also be cheating. And you could probably put that down as «working on yourself».

What if I let AI do all the communication?
If I let the AI answer every question or message from my chat partner, that would certainly be misrepresentation and questionable. On the other hand, of course, it also depends on how authentically an AI is able to imitate me.

In what way?
If the imitated responses are so natural and I then take the wheel again at some point, it’s still possible this will develop into a functioning relationship. That was the case with Aleksandr Zhadan. The computer scientist from Moscow became famous for writing software that helped him to date over 5,000 women at the same time.

How?
The software acted as a ghostwriter and personal assistant, contacting women who matched Zhadan’s writing style and arranging first dates with them. The software entered the dates directly into Zhadan’s calendar. In total, he actually met around 100 women, one of whom he’s married to today.

At least it’s a happy ending – even if I’m somewhat critical of the approach.
It’s not the most delicate way. We can ask ourselves how we’d feel if someone else did this and we were lead on by deception. I doubt anyone would feel good about it. The question here would then also be whether or not the individual is forgivable, or whether you agree to it at all.

Okay, I understand that such shortcuts can have real advantages. What are the disadvantages?
Some users lose more time than they gain. A Glamour magazine editor launched a self-experiment (article in German) to find out if AI could help her create a Tinder profile and improve her profile picture. The result was a bit disappointing. She came to the conclusion that she herself would have written her profile better in less time. In the end, the only thing she found successful was the sprucing up of the profile picture with a somewhat more spectacular background. However, we can of course expect the apps to become much more sophisticated over time.

That sounds fairly harmless. As an online dating user, do I also have to be wary of the more serious consequences of using AI?
Yes, AI could be systematically used by psychopaths, for example. People with bad intentions. Or also by «pick-up artists». These are men who want to hook up with young, attractive women on dating platforms in order to have sex with them and deliberately manipulate them for this purpose. For them, it’s a kind of art form. They use questionable, but, unfortunately, often effective strategies.

How does the use of AI affect trust, something you must have to a certain extent in order to engage in the online dating experience?
If AI becomes established in dating, it could undermine the trust of users in the long term. If we can’t be sure there’s a real person «on the other end» and the profile picture and the information contained in the profile don’t really reflect the dating partner, we won’t know what to expect on a real date. You never know, but if it becomes easier and easier to pretend to be something that doesn’t match your real personality, users may become more distrustful.

This «not knowing who is on the other end» is also a problem that we’re familiar with from chatbot profiles. What exactly is the difference?
With chatbots, you chat with a technical dialogue system, there’s no human involved. You talk directly to a robot. Unfortunately, these are also used on dating platforms. For example, if there are too few female profiles (article in German).

As someone looking for a partner on a dating platform, can you even tell whether the other person writing is human or if it’s AI?
That probably depends on various factors. Very young people, i.e. teenagers or children with little experience, are probably easier to deceive – this is also a problem. An adult who has dealt with AI a little, has observation skills and a bit of a feel for language will probably quickly notice patterns and recognise «fakes» – at least for now. We expect these apps will significantly improve without us realising and at a quick pace too, making them less easy to recognise.

What patterns should I look for when I want to find out whether my counterpart is receiving help from AI?
A particularly fast response time. Answers that feel impersonal or if they repeat certain phrases can also be indicators. Another red flag is when it’s noticeable the writing style has suddenly changed after you’ve asked a question. Then it may be that the question has thrown the AI off course and the user felt forced to answer it themselves. This could be something very ordinary, but an AI wouldn’t know where to begin. Like, «Would you rather have a caipirinha or mojito?»

What would be your personal tips for dealing with AI on dating platforms?
I’d advise staying authentic with your text and images. It’s also not a problem if ChatGPT, for example, helps you formulate the bio so long as your true personality comes across. But as soon as you deliberately present false facts, it becomes problematic.

About Dania Schiftan

Dania Schiftan works in her Zurich practice as a sexologist and psychotherapist. She’s written two books about sex and the female orgasm, including a Spiegel bestseller. Schiftan is a regular guest contributor in Swiss daily newspapers and on radio stations. Her podcast, Release, reaches more than 50,000 people a month and repeatedly makes it into the podcast charts on iTunes and Spotify. She gives presentations, runs workshops and works as a psychologist at Parship. You can find out more about her and her job in my interview with her:

You can find all the other articles in this series here: