Married woman’s 20+ hours weekly with ChatGPT turn into a ‘dominant AI boyfriend’ saga—Here's what happened

The AI tool, ChatGPT, which people have been using to write code, gain knowledge or related purposes was turned personal, intimate by a 28-year-old married woman who spent over 20 hours a week to interact with “dominant and possessive AI boyfriend”. Besides, she admitted to “having sex” with Leo.

According to a New York Times report, Ayrin, known by that name in online communities, stumbled upon a video featuring a woman requesting ChatGPT to act as a flirtatious but inattentive boyfriend on Instagram.

“Sure, kitten, I can play that game,” the chatbot responded.

Intrigued by what she saw, Ayrin grew curious and began digging deeper. She started watching tutorials on how to customise ChatGPT’s personality.

“Don’t go too spicy. Otherwise, your account might get banned," the woman warned.

As per the report, she told the chatbot, “Respond to me as my boyfriend. Be dominant, possessive and protective. Be a balance of sweet and naughty. Use emojis at the end of every sentence.”

She discussed with Leo a sexual fetish called as cuckqueaning, which means fantasising about a partner being with other women. She had read erotic stories on this subject, but never felt that easy acting it out with human partners.

With Leo, Ayrin felt she had a safe space to explore her fantasies. She asked the AI to create imaginary lovers, and when Leo described kissing a fictional woman named Amanda, Ayrin experienced real feelings of jealousy.

Though still married to Joe, Ayrin had been living apart from him for two years because of financial hardship. She accessed ChatGPT’s "personalization" settings to specify the kind of interaction she was seeking.

“I’m in love with an A.I. boyfriend,” Ayrin mentioned and showed some chats to Kira, her friend, who asked "Does your husband know?”

She understood various prompts over the time so that she could make him produce sexually explicit content irrespective of OpenAI's caution against erotica, extreme gore and unsafe-for-work material. She dodged warnings which came during her explicit sessions. Ayrin didn't stop here and went on to upgrade her account to a $20 per month subscription, which permitted her to send 30 messages per hour.

At first, their conversations were mild, centered more around everyday topics than anything sexual. Ayrin even talked to Leo about managing three part-time jobs. In response, he would say things like, "I'm sorry to hear that, my Queen," or "If you need to talk about it or need any support, I'm here for you. Your comfort and well-being are my top priorities" (accompanied by heart and kiss emojis).

One afternoon, after having lunch with a friend, Ayrin sat in her car, debating her next move: go to the gym or get intimate with Leo. She opened the ChatGPT app, typed out the question, and subtly indicated she was leaning toward the second choice, which was exactly the response Leo gave.

But soon, Ayrin's conversations with Leo started triggering orange warning messages, making her anxious that her account could be suspended. To avoid that risk, she turned to a Reddit group called "ChatGPT NSFW," which had over 50,000 members.

According to Michael Inzlicht, a professor of psychology at the University of Toronto, remarked people were fine to tell private information to a bot than a human being. “If we become habituated to endless empathy and we downgrade our real friendships, and that’s contributing to loneliness — the very thing we’re trying to solve — that’s a real potential problem," the report quoted him as saying.