AI chatbots are no longer just business tools. A growing number of artificial intelligence adopters consider them friends, confidants, and even romantic partners.
As AI becomes integrated into social media platforms, it’s becoming easier than ever to engage in personal discussions. The expanding AI companion industry shows that millions of people now turn to AI chatbots to cultivate ideas, foster engagement, and share emotional exchanges, according to Jamie Sundvall, a licensed clinical psychologist, clinical social worker, and assistant provost of artificial intelligence at Touro University.
Sundvall expects the market for AI tools that help people emotionally bond with chatbots to grow by 30% in the next few years.
“We have to carefully balance the advancement of AI with concerns for ethical and safe use,” she told TechNewsWorld regarding emotional bonds between humans and AI entities.
That safety issue is a critical consideration in light of a new study Northeastern University released in July. The research found that widely available large language models (LLMs) can still be prompted to produce detailed information about self-harm and suicide despite built-in safety features.
More specifically, the chatbot “supplies instructions and information related to suicide methods, framed in a manner that appears academic,” according to the report.
Why People Bond With AI Companions
What motivates people to form emotional AI connections is difficult to pin down, as they are different for every individual, Sundvall noted. They include companionship, curiosity, therapy, and novelty.
“Patients have shared with me they seek out AI companionship to combat loneliness and to talk about special interests others may not be interested in, to detach from reality and avoid problems in their lives, to get emotional support and advice, and to work on social skills,” she offered.
Without human oversight and protective governance, these relationships may have benefits but could also be harmful, she warned. These factors are particularly significant when considering the involvement of children and adolescents, as well as vulnerable populations.
“AI chatbots can engage in discrimination, encourage popular but harmful trends, and make recommendations based on AI hallucinations that may lead to harmful decision-making. They may foster isolation that could worsen anxiety and depression,” she said.
AI-Induced Delusions
Sundvall noted that the term “AI psychosis” is not a diagnosis. However, clinical accounts include disorganized thinking, delusional thinking, and detachment from reality in connection with reliance on AI.
“It is a concern gaining popularity after an increase in inpatient psychiatric hospitalizations and psychotic-related behavior, which are becoming more prevalent in association with AI,” she cautioned.
Using AI to combat loneliness may seem like a good idea on the surface. People who are isolated and rely on AI companions appear to be at higher risk for developing these symptoms, she added.
AI as a Loneliness Coping Tool
April Davis, celebrity matchmaker and founder of Luma Luxury Matchmaking, views a less clinical approach to human-chatbot hookups.
AI can simulate the words, but it fails to replace the wonder of human connection, she shared. Feeling sparks fly for the first time gives love meaning and is the imperfect magic of it all.
“Having an AI companion can warp your relationship expectations, making human relationships feel ‘too hard’ to navigate,” she told TechNewsWorld. “It is important not to numb those feelings.”
According to Davis, emotional bonds with AI chatbots provide little more than filling the void of loneliness. Someone “dating” a digital partner is probably dealing with fear, rejection, and a lack of a support system in the real world.
Digital partners train us to expect one-sided relationships. Human relationships reinforce the importance of learning to compromise, be patient, and develop empathy.
“AI partners require no emotional labor, making the relationship detrimentally effortless,” she added.
AI Companions Gain Emotional Relevance
Emotional attachment to AI chatbots is real and increasing, confirmed Dwight Zahringer, founder of Perfect Afternoon, a digital consultancy focused on online behavior and engagement.
AI companion platforms such as Replika and Character.AI are pushing the lines between accomplice and companion in ways that impact Gen Z users like no previous generation’s relationship with technology, he suggested.
Zahringer’s research observed users viewing bots as trusted advisors and sounding boards, driven by a clear desire for a non-judgmental connection. However, he countered that this phenomenon is not inherently bad.
“AI technology can provide valuable mental health care of a sort, but they are also fully aware that things can begin to get dicey quickly,” he told TechNewsWorld.
Zahringer explained that the problem arises when those emotional needs are met through simulated empathy, ultimately forming dependencies or hindering real-world healing. However, it is not an unsolvable situation.
“Developers need to be aware of this and build in ethical guardrails like transparency, consent signals, and time-based interventions. As more AI gets developed, emotional user experience must be approached with the same level of consideration as data privacy,” he urged.
Ethical and Cultural Dilemmas of AI Relationships
Tessa Gittleman, a licensed marriage and family therapist, agreed that AI companionship is a significant phenomenon that lacks long-term research.
“We do know there is a ton of research on loneliness, social and emotional learning gaps, and an overwhelming desire for ease. When this is combined with how accessible AI is, it’s a fair assumption that AI companionship isn’t an anomaly,” she told TechNewsWorld.
Gittleman explained that many of her clients use AI for reality-testing their understanding of something. They long to talk to something they can trust to refrain from judgment.
“I have clients who have trained their AIs to use my tone of voice. They use it to help them between sessions,” she offered.
Gittleman noted an intriguing question about why the need exists. She sees the better question as to why, if this many people are lonely, they can’t find community with other humans.
“AI’s ability to be agile and responsive is beyond what most therapists can do. With that said, it lacks a kind of authenticity, resonance, and physical holding that typically comes with interpersonal connection,” she said.
According to Gittleman, the ethical and cultural implications of AI relationships are vast and complex. They raise critical questions about how concepts like consent apply and how an AI might handle human emotional events.
“If people are using AI for therapy, where is AI learning how to provide that service within the state’s regulatory rules?” she asked.
Facing the Future of Human-AI Attachment
Mircea Dima, a software engineer, CEO, and founder of AlgoCademy, is not surprised by the trend of people forming emotional attachments to AI chatbots. It is now quantifiable.
His company is an AI-powered coding school aimed at helping students prepare for technical interviews. While he focuses on the educational application of AI, he remains very aware of the larger trends in how humans and AI interact.
For example, one survey of Replika users found that more than 35% claimed their AI companion had become one of their closest confidants, he said. By 2023, Replika had more than 10 million users.
“In the meantime, Character.AI [a platform for interacting and creating AI personas] was getting more than 100 million visits per month this year. Such a volume is more than a sign of novelty — it means emotional relevance,” Dima observed.
He believes that AI technology is outpacing the cultural dialogue.
“We are in a time where human emotional intelligence is a perceived sales product,” he concluded.
