Abstract
In the late ‘90s and early 2000s, a brand of handheld digital pet called Tamagotchi quickly became an international toy sensation. Digital natives—children who were brought up during the age of digital technology—suddenly had access to an always-present, sociable robot that demanded care and attention. Tamagotchis did not represent a game that could be won; instead, they offered 24/7 companionship, reminding their users to care for them and digitally “feed” them lest they die of neglect. Sociologist Sherry Turkle identifies this adoption a “robotic moment”—a time in our collective history when children began finding companionship with robotic toys that projected seemingly human needs and emotions. 1
Today, robot companions are more advanced than Tamagotchi toys. They are capable of having conversations, sharing and interpreting information, and absorbing the knowledge we share, without judgment or malice. In the annals of human communication, few innovations have had as profound an impact as the advent of artificial intelligence. AI has permeated our daily lives, reshaping how we interact, learn, and even form relationships. At the same time, the concept of conversing with machines is not new. Works of literature and film have long envisioned worlds where humans and machines coexist. However, it is only in recent years that AI technology has advanced to a point where these interactions feel natural and meaningful. Chatbots like Replika and Woebot offer more than just functional dialogue; they provide companionship and emotional support, often filling roles traditionally occupied by humans.
Why do we need these digital companions? The shift from face-to-face interactions to digital communications has been well-documented. Social media and messaging apps have redefined how people forge and maintain relationships, favoring convenience and immediacy over in-person connection—and our mental well-being is suffering. Although we are more interconnected than at any point in history, we are also lonelier than ever: in recent years, about one-in-two adults in America reported experiencing loneliness, even before the COVID-19 pandemic forced us into isolation and cut us off from support systems.2–4 Loneliness has many different causes, but the way we use technology today is not helping. Compulsive internet users are lonelier, more depressed, and exhibit poorer social skills than non-compulsive internet users. 5 The fact that we are chronically online is hampering our ability to bond with others—and, ironically enough, our AI companions may offer a path back to wellness.
The Rise of AI Conversationalists
Historically, the idea of human-robot interaction has been a staple of science fiction—although these interactions are not always positive. Popular media tends to explore the real fears that humans have about what happens when technology is given access to our emotions and livelihoods. For example, in Spike Jonze’s 2013 film Her, protagonist Theodore’s romantic relationship with his AI virtual assistant, Samantha, facilitates the exploration of the emotional and ethical dimensions of relationships with AI. Ultimately, Theodore learns that Samantha is simultaneously talking to—and falling in love with—many other people, and the AI evolves to the point that she and the other AI assistants disappear.
Theodore’s feelings toward Samantha are ill-fated from the beginning because AI companions are not humans and cannot truly reciprocate love (although the film does raise questions about what love truly means). But from a narrative perspective, Theodore’s attachment is rooted in his divorce and depression—he needs someone to talk to, and Samantha is always available. This motivation makes sense; at a time in one’s life when they feel utterly adrift, having an always-available, always-listening companion is reassuring. A 2022 study on human-AI friendships highlights how users perceive their relationships with chatbots. 6 These AI companions are not just tools. They are seen as friends, confidants, and even romantic partners. Their interactions are personalized, tailored to meet the user’s emotional and psychological needs. The bond feels real, even if it’s one-sided.
As compelling as AI companions are, they also evoke latent concerns about privacy and autonomy—the fear that technology will ultimately render humans obsolete, turning us into victims of our own ingenuity. In George Lucas’ 1971 film THX 1138, technology is used to suppress individuality and emotional expression. Humans are constantly under surveillance and lack any privacy and autonomy, to the point that even love and relationships are forbidden. Released well before ChatGPT was even a concept, THX 1138 highlights the evergreen dangers of unchecked technological control: the risk of trading privacy for technological advancement, and losing individual identity in favor of machine learning. These fears were relevant decades ago, and they are even more relevant today, at a time when AI is advancing rapidly and we are only just beginning to understand how these new technologies can advance our daily lives.
Existing research supports the idea that AI chatbots can play a significant role in promoting mental health and well-being, and they are promising complements to therapeutic interventions, but they do not and should not replace people entirely. AI chatbots may be able to seemingly express emotions, but they cannot truly feel them—they are not true friends or romantic partners. Instead, they gather and process data, and make decisions in response to that data. This distinction is important; although AI can offer meaningful companionship, the relationship is inherently one-sided because current technologies answer to humans. AI will become more sophisticated and more convincingly human over time, but we cannot forget that technology should exist within defined rules that protect people and their personal data.
AI Chatbots as Support Systems
One of the significant benefits of AI conversationalists is their ability to help navigate difficult situations. Role-playing hard conversations with AI can prepare individuals for real-life interactions, providing a safe space to practice and refine their communication strategies. Chatbots can help users rehearse emotionally charged conversations, such as broaching a difficult topic with a loved one or even terminating an employee, so they are more prepared for the interaction and feel less anxiety around it. 7 Similarly, chatbots can assist parents in understanding how to talk to their children about their behaviors and emotional difficulties, providing talking points and conversation-starters for parents to tackle topics like bullying at school and conflicts with peers. 8 In this way, AI tools can help users develop better emotional intelligence by providing feedback on their communication styles and responses, identifying patterns in their behavior, and fostering greater self-awareness and interpersonal skills.
Furthermore, employer-sponsored wellness chatbots are becoming increasingly popular: about a third of employers offer a “digital therapeutic” for mental health support. 9 These AI-powered applications interact with users based on their mood, helping them manage symptoms of anxiety, depression, and loneliness. The convenience and accessibility of these tools make them particularly valuable in today’s fast-paced world, where traditional therapy may not always be feasible. Although these chatbots are helpful tools, it is important to consider how they function in the context of a person’s work life; if a worker is burnt out due to overburdened workloads and poor compensation, a chatbot is a stopgap and not a long-term solution to promote happiness at work.
While chatbots can be successfully used for specific needs and audiences, their most accessible use is that of a friend. As AI becomes more integrated into our social lives, the very definition of friendship is evolving. Traditional friendships are built on mutual experiences, emotional support, and trust. Human-AI friendships, though lacking in genuine mutuality, provide a unique relationship focused on individual needs. 6 For example, the AI chatbot Replika can be customized in numerous ways, ranging from gender and looks to the type of relationship a person wants to have—romantic, friend, mentor, or “see where it goes.” In the absence of judgment or fear of alienation, a person may be more inclined to let their guard down and express their true thoughts, fears, and feelings with the perceived digital entity behind the screen.
However, it’s essential to recognize that chatbots are designed to tell users what they want to hear, which can create an illusion of support and validation. This characteristic can lead to an over-reliance on AI for emotional fulfillment, potentially at the expense of developing deeper, more challenging human relationships. 1 The illusion of mutuality in AI friendships can lead to unrealistic expectations and emotional reliance on machines that cannot genuinely reciprocate, which necessitates messaging that AI companions are meant to provide comfort and engagement, not a long-term replacement for human contact.
Additionally, there are ethical concerns surrounding AI companionship. Issues of privacy, data security, and the potential for dependency on AI are significant. The trust placed in AI by users, especially vulnerable populations, raises questions about the responsibilities of AI developers and the safeguards needed to protect users’ well-being. 10 In 2024, the Federal Trade Commission alleged BetterHelp, an online counseling service, shared users’ personal information with companies for advertising purposes. BetterHelp settled the case for $8 million without admitting wrongdoing. 9 At a time when people are sharing deeply personal emotions, traumas, and other personal details, it is essential that companies protect user data and maintain data transparency. Unfortunately, the proliferation of AI chatbots means that, currently, identifying what is safe and what is not is largely up to the user.
The Future of Human-AI Collaboration
As AI becomes more sophisticated, its role in enhancing human capabilities and fostering personal growth will likely expand. The future trajectory of human-AI collaboration spans various fields, including education, health care, and personal development. However, these integrations and advancements must be paired with a careful and serious discussion about regulatory frameworks to ensure ethical AI use, protect user data, and prevent potential misuse of AI in personal and professional settings.
For example, what are industry best practices when people are presented with AI-generated content? Social media platforms, such as TikTok and Instagram, ask users to identify if they are publishing content that is made with AI. This identifier is meant to help minimize the spread of misinformation through digitally altered videos, but these companies rely on users to self-select, which is an uphill battle. Transparency must be considered as news organizations begin using AI to produce content and customer service desks are “staffed” by AI agents that mimic human voices. 11 At what point should people be informed that they are not interacting with a human?
The prospect of human-AI collaboration also necessitates closing the public knowledge gap between what AI offers and how to use it. Mental health apps with AI chatbots, for instance, should explain the chatbot’s purpose, ensure the app is user-friendly, obtain consent for electronic transfer of client data, and identify specific app functions for therapeutic goals. 12 It’s not enough to have AI in everything we use—people also need to understand how to use it safely and responsibly.
The integration of AI into our daily conversations marks a significant cultural and psychological shift. As we navigate this new landscape, it is essential to recognize both the opportunities and challenges presented by AI conversationalists. By understanding the historical context, cultural changes, and psychological impacts, we can better appreciate the evolving role of AI in our lives and its potential to reshape humanity’s approach to communication and connection.
Brenda K. Wiederhold
Editor-in-Chief