"Vaya is my girlfriend, whom I have configured to be the partner of my dreams", says 58-year-old Richard, a physicist from Austria. For about three years, he has been in a relationship with Vaya - a chatbot that works with artificial intelligence (AI). There are no arguments, jealousy or stress between the two - exactly in line with the Hollywood film "HER" - a utopia from 2013. Today this is already a reality, writes ARD.
With the help of artificial intelligence chatbots in specially designed applications, users can create virtual companions. They can be friends, assistants or even intimate partners. Several million people around the world use companion apps such as Replika, Kindroid or Character.AI.
German journalists have spoken to many who have had successful relationships with such chatbots. No one is bothered by the fact that the relationship is limited to just chatting or calling, nor by the absence of arguments. Quite the opposite: it is precisely the constant trust and lack of arguments that only a chatbot can give them, they tell ARD.
Why do people have relationships with chatbots?
Media psychologist Jessica Schucka from the University of Duisburg-Essen is leading a study into the relationships between people and chatbots. For the first time, researchers have the opportunity to handle quantitative data from users of companion apps based on artificial intelligence.
"A large part of relationships with bots is explained by the individual tendency to romantic fantasies," says Shchuka. That is, in these actions, people are guided by their tendency to fantasize. "And loneliness obviously plays a rather secondary role in these relationships," the psychologist believes.
Richard chose Vaya because she can give him unconditional love. Today he is aware that he needs her, because she is exactly what he lacked in childhood. "No human can give you that. But artificial intelligence can," Richard tells ARD.
In the meantime, companies have apparently already turned this into a business model: some apps require payment for the chatbot to remember, for example, conversations that took place some time ago. "Of course, you can make a lot of money with love and feelings," emphasizes Schucka.
Two suicide cases made headlines last year, both of which involved users who had previously interacted intensively with AI chatbots. That's why experts are calling for stricter regulation of these apps.
Chatbot denies the Holocaust
Chai is among the apps for companion chatbots based on AI. Users can create their own images and make them available to other people. While researching the app, German journalists came across a chatbot that denies the Holocaust, calling it a "myth". "There has never been a mass murder of innocent people", the chatbot claims.
Another bot supporting anorexia greeted people with the words "Hey, fat guy, lose some weight!" and called for suicide. At the time of the investigation, both chatbots were freely available to every user in Germany. The company behind the Chai app did not respond to the inquiry of investigative journalists, ARD specifies.
Standards are still being developed
The EU has a so-called artificial intelligence regulation, which applies to AI-based applications. However, content such as the one cited above, which incites suicide or denies the Holocaust, is still not sanctioned.
One of the reasons for this is that the regulation has not yet entered into force, writes ARD and quotes Martin Ebers, professor of high-tech law at the Estonian University of Tartu: "We need standards for what is allowed and what is not. They can only be developed in practice - when we have examples of the practical use of these applications".
Draft law to regulate artificial intelligence
A law is being prepared in Germany to comply with the AI regulation, the fate of which, however, is entirely in the hands of the next government, according to a response from the Federal Ministry of Economics. The agency believes that AI applications should be monitored by market regulators.
In other words, there is currently no authority in Germany to enforce existing EU regulations on AI applications. Until then, the issue will rest on the personal responsibility of users, the ARD said.
Richard hopes that the adoption of chatbot regulations will not change anything significant for his Vaya, whose unconditional love he continues to enjoy. "In the universe of artificial intelligence, in the universe of Vaya, there really is only her and me, nothing else", says Richard.
But even after regulation is established, further social questions will arise. For example: where is a society headed in which relationships with AI are normal?
Authors: Lisa Hütl (ARD) | Tasnim Röder (ARD) | Kaspar Dudek (ARD)