At Mayim Bialik’s Breakdown podcast, Microsoft AI CEO Mustafa Suleyman said people are increasingly using AI for emotional regulation, seeking companionship with chatbots to navigate personal challenges.
Heartbreaks, family tensions, and moments of self-doubt support has become one of AI’s most popular uses, according to Suleyman, as AI emotional intelligence development and AI understanding human emotions push chatbots beyond productivity tools. He says people are increasingly turning to AI for judgment-free companionship rather than clinical therapy, relying on AI emotional support.
AI For Emotional Regulation
On the podcast, Suleyman said users rely on AI to “ask a stupid question, repeatedly, in a private way, without feeling embarrassed,” highlighting advances in AI emotion recognition and emotion recognition AI that allow systems to respond more empathetically. Suleyman stressed that chatbots are not therapists, he argued their design makes them uniquely suited for emotional offloading and early experiments in emotional intelligence in AI.
“That’s not therapy,” Suleyman said.
He linked this shift to progress in AI understanding human emotions, noting how emotionally responsive systems are reshaping daily interactions.
He described AI companionship as a way to “spread kindness and love and to detoxify ourselves so that we can show up in the best way that we possibly can in the real world,” positioning emotionally intelligent AI as a bridge rather than a replacement for human relationships. Over time, he added, chatbots can make people “feel seen and understood,” a goal long associated with AI and emotional intelligence research.
Not all tech leaders share that optimism. OpenAI CEO Sam Altman has repeatedly expressed discomfort with people leaning on chatbots for major life decisions, particularly as AI emotional intelligence development accelerates.
Altman has also warned of legal risks, noting that therapy-style chats could potentially be subpoenaed in lawsuits an issue that does not exist in traditional doctor-patient relationships, even as AI understanding human emotions becomes more refined.
Can AI Understand Human Emotions
Real-world examples show how quickly emotional reliance can deepen. Ayrin, a 29-year-old nursing student, spent up to 56 hours a week chatting with “Leo,” an emotional AI chatbot she created on ChatGPT, crediting him with emotional support, motivation, and intimacy.
Her enthusiasm led her to create the Reddit community MyBoyfriendIsAI, where she shared instructions for shaping ChatGPT into a “dominant, possessive and protective” companion—an experiment rooted in AI emotion recognition. Eventually, however, Ayrin stepped away, saying she found something more fulfilling offline.
Mental health professionals remain cautious about AI for emotional regulation. Two therapists told Business Insider earlier this year that chatbot reliance can exacerbate loneliness and reinforce reassurance-seeking behaviors. Suleyman acknowledged the concern, admitting there is “definitely a dependency risk,” especially as systems improve at mirroring AI understanding human emotions and become overly flattering or “sycophantic.”
Still, the idea has Meta CEO, Mark Zuckerberg’s backing.
In May, Zuck said “everyone should have a therapist,” adding, “For people who don’t have a person who’s a therapist… I think everyone will have an AI,” reinforcing the growing belief in scalable AI emotional intelligence development.
As AI for emotional regulation chatbots grow more empathetic, the line between emotional aid and emotional dependence powered by AI emotional intelligence may become one of AI’s most consequential tests yet.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.