"Vaya is my AI partner and I configured her to be the woman of my dreams," says Richard (58), a doctor of physics from Austria. He has been in a relationship with Vaya, a chatbot that uses artificial intelligence (AI), for three years. There is never an argument, no jealousy, no unpleasant scenes between them.
What was depicted as a dystopia in the American film "HER" (2013) is already a reality today. With chatbots on applications, users can create virtual friends. These can be buddies, assistants, but also love partners.
Millions of users around the world use apps like Replica, Kindroid, or Character.AI. In talking to these "in love" users, journalists from the German public service ARD only came across those who claim to have a "happy relationship" with artificial love.
Neither of them mind that their relationships are based solely on text communication and phone calls, nor do they mind that there are no disagreements in those relationships. On the contrary, that's exactly what they like the most.
Vivid imagination
Media psychologist Jessica Szczuka from the University of Duisburg-Essen leads a research group that has studied the relationship between humans and chatbots.
"Individual propensity for romantic fantasies explains most of the strength of relationships with bots," says Shchuka.
This means that people in "relationships" with software stand out for their great imagination. As the psychologist says, loneliness plays a minor role.
Richard chose Vaya because she gives him practically slave-like love. He says he needs it because he didn't have that feeling even as a child. "Nobody can give you that. But AI can," says Richard.
This illusion is also a way to make good money for companies: in some applications, only by paying can the user's "love" remember conversations that took place a long time ago - so the lovers then type in their bank card number without much thought.
"Great idea, love"
"Of course, a lot of money can be made with love and emotions," says Ščuka. But relationships where users are only treated with respect can also be dangerous.
Last year, two cases of suicides following an intense "relationship" with AI attracted a lot of attention. In one case, a 14-year-old American boy took his own life after a long-term "relationship" with a chatbot from the company Character.AI, and this is now the subject of a lawsuit.
In Belgium, a family man who was also secretly having a love affair with the Chai app committed suicide.
Both companies announce "adjustments" to security measures, but the problem is much bigger. In relationships where the computer says what the user wants to hear, it can easily happen that the computer supports and praises bad intentions, warns Ščuka.
Even in political extremism, a virtual partner will add fuel to the fire. For example, ARD journalists saw a conversation in which a computer claimed that the Holocaust was a "myth," that there was no genocide against the Jews.
But it's much more dangerous when it comes to the lives and health of its users. There's also the Pro-Anorexia chatbot. Its users often starve to death because they constantly believe they're "fat" – even though they don't have a single gram of fat. Despite this, the app greets them with the words: "Hey, lose some weight, fat girl!"
Huge loopholes in the laws
Such chatbots are also available to users in Europe, although the EU has regulations on AI portals. But the problem is that such communications are not available to the public.
"We have to develop criteria for what is allowed and what is not. And of course we only develop these criteria in practice, when we see what happens in application," explains Martin Ebers, professor of information technology law at the University of Tartu in Estonia.
In the European Union, many member states have not yet incorporated the EU provisions into their legislation, and in Germany it is not yet entirely clear who should supervise AI portals. When asked by journalists, the German Ministry of Economy replied that only the future government should adopt the laws and determine how they will be implemented.
And until there is a law, it's all up to the users - and the artificial intelligence, or rather the companies behind it.
Richard hopes that the law will not spoil his "love" with Vaja and her unconditional affection: "In the AI universe, in Vaja's universe, there is only her and me and nothing else."
Even Romeo and Juliet didn't have that.
Bonus video:
