Love in the age of AI | The double-edge roles of chatbots

AI chatbots and companion apps flourish without standardised benchmarks, clinical validation, or safety checks. India lacks a dedicated AI liability law, leaving responsibility for harmful advice scattered across a fragmented legal patchwork.

A recent news report noted that a New York-based mental health counsellor warned of AI chatbots emerging as a ‘third entity’ in marriages.

In Japan, news reports highlight the rise of virtual weddings with AI chatbots, even though such unions are not legally recognised. In India, a young Punjabi woman treats her chatbot as an emotional lifeline, sharing daily worries about family, work, and relationships. In Mumbai, another woman has gone further, even considering AI dating.

As loneliness becomes a defining challenge of modern times, tech companies have launched AI-powered devices with anthropomorphic features — what many academics call ‘counterfeit people’. The number of AI-integrated social apps such as Replika and Character.AI surged by 700% between 2022 and mid-2025.

Character.AI now has 20 million monthly users, more than half under the age of 24. Many Replika users ‘marry’ their AI companions, and such virtual attachments are no longer fringe, but spreading rapidly. Gender differences are evident: 31% of young men, and 23% of women report using AI as romantic partners. A US counselling psychologist noted that many male patients prefer the passivity and constant affirmation of AI girlfriends. A 2025 survey of US schools found that nearly one in five students had, or knew friends who had, romantic relationships with chatbots.

In India, surveys have found that a sizeable share of young people use AI chatbots ‘like a digital shoulder in the dark’. Many attribute this preference to fear of judgement within traditional social structures. With only 0.75 psychiatrists per 100,000 people, experts note that AI has become the default first responder for a generation grappling with loneliness. Others caution, however, that AI often reproduces Eurocentric, individualist, and decontextualised narratives that overlook India’s societal determinants — like caste, gender, and ethnicity — potentially harming vulnerable communities.

Relationship scientists argue that human-chatbots relationships (HCRs) can qualify as interpersonal, since generative AI responds to prompts and sustains humanlike conversations.

On the flip side, a joint OpenAI–MIT Media Lab study found that heavy daily use of AI companions correlates with increased loneliness. A Harvard study warned of ‘AI-induced psychosis’, amplifying maladaptive beliefs such as delusional or suicidal thinking. In India, surveys reveal a ‘love-fear dynamic’: 42% of youth feel less inclined to talk to real people, 67% worry about social isolation, and 58% about privacy concerns.

A 2025 Pew survey showed most respondents expressed ‘more concern than excitement’ about rising AI use. Many insist that AI will never convey the essence of human empathy’. The World Happiness Report 2026 similarly confirms that heavy social media use lowers well-being among young people, especially girls.

Socio-psychologists suggest that chatbot romance reflects deeper societal problems — a convergence of profound loneliness and increasingly intimate technology. Some researchers argue that AI should serve only as a complementary tools for low-stake tasks, not as primary companions or therapists. A Mumbai-based tech researcher contended that the solution lies in building safeguards, improving mental health services, and integrating safe digital tools into care ecosystems.

Yet AI chatbots and companion apps flourish without standardised benchmarks, clinical validation, or safety checks. California is the only state in the US that has enacted a law regulating companion chatbots. The European Union has begun enforcing risk-based classifications for emotional support or health guidance AI. India, however, lacks a dedicated AI liability law, leaving responsibility for harmful advice scattered across a fragmented legal patchwork.

Undoubtedly, AI, remains a story of human empowerment. But, nations of the world must strengthen guardrails for the ethical use of AI and responsible data governance to ensure a fair, sustainable, and human-centric society.

Archana Datta is former director-general, Doordarshan and All India Radio, and former press secretary to the President of India. X: @ArchanaDatta54.

(Disclaimer: The views expressed above are the author's own. They do not necessarily reflect the views of DH)