AI companions and the loneliness epidemic: The importance of human connection in a digital era

As “friend” AI ads take over the subway, they raise an unsettling question: are we outsourcing connection to machines? 

Artwork by Vernon Demir

By Katie Denissen

November 2, 2025

Minimalist black-and-white text, an image of a large pearl-like pendant. A definition:

friend

[frend] noun

someone who listens, responds, and supports you.  

Train car cards with large promises plastered across a white backdrop:

I’ll never leave dirty dishes in the sink.

I’ll never bail on our dinner plans. 

Though, there’s more to this picture. If you take the subway in NYC, you most likely have seen the controversial advertisements that have taken over train cars and station walls: Friend, the new wearable A.I. pendant that functions by listening to conversations and acting as a digital companion. These posters — part of the startup’s one million dollar ad campaign — have sparked a pretty intense reaction, as New Yorkers took to vandalizing these simply-designed posters. Scrawled marker defaces the clean lines of the advertisement with expressions of outrage and rejection:

“Stop profiting off of loneliness”

“Talk to real people”

“AI Trash”

“Sureveillance capitalism, get real friends”

Photo by Katie Denissen/The Barnard Bulletin

Messages like these, which now seem to be found on every advertisement, reflect growing societal unease in regards to AI. Current opinions seem to be critical as developments continue to roll out, and the future of artificial intelligence in the social realm is still uncertain. 

Barnard Sociology professor Bizaa Ali says the backlash was predictable. “Very frankly, there is nothing novel or innovative about the concept of an AI companion — there are many in the market already,” she explains. “What was significant in this case was the scale of the marketing for this product and the tone-deaf messaging in the ads, which predictably generated what we can call the ‘Anti-Friend backlash.’  This necklace clearly is a useless luxury surveillance product with multiple privacy and transparency issues, and people recognize this.”

This negative reaction seems to be a sentiment shared by other members of the Barnard community. “When I see the ads I get extremely irritated and worried about the future of AI,” says Barnard student Hoodo Ali (BC ’28). “I think friend.com AI is profiting off of people’s loneliness.” 

According to Professor Ali, the controversy surrounding Friend AI was not entirely unintentional. “While the Friend AI advertisements come across as a marketing disaster, they were apparently meant to be provocative,” she explains, “as a sort of click-bait or ad-baiting strategy. We can describe this as a socio-technical experiment.”

Along with the complicated issue of the AI technology itself,  the definition of “friend” that the advertisement offers is also deeply flawed.  Defining “friend” as “someone who listens, responds, and supports you” frames friendship purely in terms of what someone does for you. It reduces connection to a transactional experience centered on the self. The language reinforces a self-focused notion of companionship, implying that a “friend” exists to meet your emotional needs rather than to share in them. This definition reflects a broader cultural pattern in which relationships are measured by personal benefit rather than by the willingness to give, forgive, and grow alongside someone else. By promoting a one-sided idea of friendship, the campaign fundamentally misunderstands human connection and transforms the benefits of relationships into short-term gratification.

Yet, as Professor Ali points out, the public response reveals something positive: “the public reaction to this campaign showed that people generally have the digital literacy and strong opinions about the techno-moral consequences of such products and are even willing to speak out publicly,” she says. “There are, of course, vulnerable demographics like children and elderly people who are often targeted by such campaigns and products and are the unwitting recipients of such ‘gifts of surveillance.’”

The backlash to Friend AI echoes earlier public resistance to intrusive tech. “Many years ago there was a similar backlash against Google Glass,” Professor Ali notes. “People came up with the term ‘Glassholes’ to describe people who use them. The terminology reflected social attitudes and was one of the reasons the product failed. While discussing the Friend AI controversy in class, many students suggested we need a similar term like ‘friendholes’  to force people to think twice before using such technology.”

This philosophy guiding the slogans displayed on these advertisements also seems to take advantage of the loneliness epidemic.  In the spring of 2023, United States Surgeon General Vivek Murthy drew national attention to America’s growing loneliness crisis by officially declaring it an epidemic. In an accompanying advisory letter, he emphasized that loneliness is more than “just a bad feeling”; it poses a serious public health threat to both individuals and society as a whole. The reason for this severity is that loneliness and isolation have been linked to serious health concerns; they have been found to increase the risk of stroke, diabetes, heart disease, cognitive decline, and premature death, and cause or intensify mental health struggles like depression, anxiety, and thoughts of self-harm or suicide.

As in human nature, we yearn for connection and companionship, which seems to be becoming harder to obtain. In fact, more recent polls have shown that as many as one in five Americans struggle with loneliness. This marks a notable rise from earlier in 2024, when loneliness hovered around 17 to 18 percent, signaling a reversal of the gradual improvement seen since the height of the pandemic. In 2021, during the early months of COVID-19, loneliness rates reached as high as 25 percent, suggesting that while we’ve made progress since those isolating days, the U.S. has not entirely recovered its sense of connection. But what is causing this crisis? Further, what can we do about it?

A Harvard survey led by researchers with Harvard Graduate School of Education’s Making Caring Common (MCC) project investigated not online the current state of loneliness in the country, but the causes behind the increase in prevalence of this feeling. Amongst the leading causes were “Technology” — 73 percent of responses — and “Living in a society that is too individualistic” — 58 percent of responses. 

While technology connects us more broadly than ever before, it also paradoxically deepens our sense of isolation. As a college student, it is a comforting feeling to be able to keep up with friends and family across the country through technology and social media. However, the constant stream of curated lives on social media and the convenience of digital communication have redefined what it means to “be connected.” Many people are surrounded by virtual interaction, yet starved for genuine intimacy. Apps, chatbots, and now AI companions promise to fill that void, but they risk replacing human presence with simulations of it. A wearable “friend” that listens and responds may sound comforting, but it raises an uncomfortable question: are we outsourcing empathy to machines?

The idea that we live in an increasingly individualistic society is another major driver of loneliness. Modern culture often glorifies self-sufficiency and independence. These ideals, while empowering in some ways, can also isolate us. This can be seen especially through self-care trends that are popular on social media; the repeatedly pushed message of “putting yourself first” encourages us to prioritize personal achievement over community, and the wants and needs of the self over altruism and kindness. Yes, taking care of yourself is indeed important and crucial; however, it seems as though this message has become over-amplified. There is a balance to be stricken. In this environment, vulnerability and dependence on others are often seen as weaknesses rather than essential parts of being human. The result is a social landscape where people are constantly surrounded by others yet feel profoundly unseen.

This hyper-individualism seeps into so many aspects of daily life, from the way we work to the way we communicate. With remote jobs and self-checkouts, many of our routines now require little to no real human contact. Even friendship is often reframed as a matter of convenience, as something to fit into an already packed schedule rather than a vital emotional bond to nurture. 

This can be a pervasive reality in college. Despite being constantly surrounded by people, many students may feel disconnected, unseen, or out of place. The transition to college often involves leaving behind familiar support systems and being thrust into an environment where everyone seems to be forming friendships effortlessly. Social media only intensifies this illusion. Academic pressure, competition, and the expectation to “make the most” of the college experience can further discourage students from admitting when they are struggling. This can lead to self-isolation. When connection becomes optional, loneliness becomes inevitable. It is no wonder that AI “companions” are emerging at this cultural moment. They promise the illusion of a relationship without the inconvenience of reciprocity.

Experts warn that although AI companionship may temporarily ease feelings of loneliness, it could reinforce the very problem it claims to solve. Human connection is reciprocal. It involves vulnerability, patience, and imperfection. A programmed companion, no matter how advanced, can only imitate empathy — it cannot truly feel it. Relying on such technology might make it easier to retreat from the effort required to form and sustain real relationships. In doing so, we risk trading authentic social bonds for convenient digital comfort.

Studies have consistently shown that strong social connections are one of the most powerful predictors of long-term health and well-being. According to the Harvard Study of Adult Development — the longest-running study on human happiness — close relationships, more than wealth or fame, are what keep people happier and healthier throughout their lives. Meaningful human relationships reduce stress, boost immune function, and even lower the likelihood of heart disease and depression. To put it simply, connection is not just good for the soul, it is also vital for survival. As humans, we are biologically and psychologically conditioned for connection, as our nervous systems are wired for empathy and our brains seek social bonding for survival and well-being.

Still, it is important to recognize why innovations like Friend are emerging now. They respond to a genuine societal need. In cities like New York, where millions live side by side yet often remain strangers, technology becomes a coping mechanism for disconnection. AI companionship, then, may not be the root of the problem but a symptom of it.

When we start to outsource companionship to machines, we risk dulling the very skills that make connection possible: listening, empathizing, forgiving, and being present. Real friendship demands effort and vulnerability, but that is also what makes it so meaningful. The appearance of the AI Friend ads that now line subway walls forces us to confront a difficult truth: loneliness cannot be solved by convenience. What we need is not a smarter device but a stronger commitment to each other.