AI Ethics & Policy

AI Companions Are Growing Alongside Real Friendships

AI companionship is common but not the main story. A Gallup study conducted with the Boston Children’s Digital Wellness Lab found that young people still rate in-person friendships as their strongest connections, even as digital relationships and chatbots become part of everyday social life.

The study surveyed 15,925 people aged 15 to 22 across 16 countries using online questionnaires conducted between May and August. Around 93% reported having at least one strong in-person friendship, and the researchers found that digital friendships usually complemented those relationships rather than replacing them.

Stronger friendships were linked with greater happiness regardless of how people connected. The report examined in-person friendships, relationships that began in real life but continued online, and online-only friendships where the people involved had never met in person.

AI is becoming social by degrees

One-third of respondents said they had an AI companion. That figure rose to 50% among daily AI users and reached 42% among young people with online-only friends, suggesting that frequent AI use and thinner social networks often appear together.

AI use varied by country. Young people in the United States, Canada, and the United Kingdom showed the least interest, with only 20% saying they used AI once a day or more. Nearly 40% of young people in Brazil used it daily, while Japan and Italy each came in around 35%.

Nearly half of young people in the United States, United Kingdom, and Canada had used AI no more than a few times in the past month, including those who had not used it at all. Across the full study, 82% had used AI to look up information, 68% had asked it for advice, and 33% had used it as a companion.

David Bickham, research director of the Digital Wellness Lab, rejected the idea that young people turning to AI automatically signals some major personal deficit. He said they are growing up with AI as “the most powerful way to find information online,” and that using it more exposes them to its social side.

“In-person and digitally mediated friendships coexist within the same social life,” the report’s authors wrote. That coexistence is not a clean victory for either side—technology has never been especially interested in clean categories.

Loneliness remains the difficult variable

Young people who used AI as a friend reported higher levels of loneliness. Those with online-only friends also showed higher loneliness than people who frequently communicated with in-person friends or family, and the pattern did not change based on how highly they rated the quality of their online-only friendships.

The researchers stressed that the findings do not establish cause and effect. They cannot determine whether lonely young people turn to AI, whether AI companionship changes their social experiences, or whether other circumstances explain both patterns.

Bickham said loneliness may be something online relationships cannot fully compensate for. Young people who feel lonely often talk to strangers online while trying to make friends, a behavior that carries risk but also reflects how many friendships begin.

The study found a similar pattern in how young people spend time online. Seventy-two percent said they spend their time scrolling short-form content, mostly on TikTok and Instagram, while six in 10 said they often or always scroll through online content when lonely.

Scrolling and spending time in person produced nearly identical ratings for perceived helpfulness as coping strategies. Yet young people with strong in-person friendships reported feeling lonely less often, giving real-world connection the one advantage that keeps surviving every app launch.

The report also connected social circumstances with access to meaningful relationships. Among young people who had none of the friendship modes examined, the strongest associated factor was lacking a trusted adult outside the family. That association remained substantial after accounting for country or residence, gender, age, urbanicity, and parent or caregiver relationship quality.

The authors warned against treating AI companionship and human friendship as interchangeable measures. Young people may expect different things from each, so comparisons require caution.

The uneasy rules of artificial intimacy

Other findings show that discomfort around AI is not limited to companionship. A Milltown Partners survey found that 73% of Gen Z adults had not made up their minds about AI, while 50% expected it to harm society. Baby boomers came in at 50%, Gen X at 48%, and millennials at 39%.

Forty-six percent said they used AI every day, and 30% of Gen Z respondents said they sometimes felt guilty using it. Experiments led by Danielle Hass at Kennesaw State University found that people who had used AI before felt just as guilty as those who had not when AI helped write personal messages.

“Everyone felt guilty,” Hass said, pointing to dishonesty and the feeling of manipulating someone as the main drivers. Scott Claessens, a senior researcher at the University of Bristol, said people want effort in social relationships because effort signals investment.

Sarah Newman of Harvard University’s metaLAB offered the less technological answer: “Personal messages don’t need to be perfect.” Their imperfections can make them authentic and endearing, which is a difficult feature for a chatbot to reproduce without turning sincerity into a setting.

The broader finding is less dramatic than the usual AI panic and more useful. The quality of young people’s relationships, their wider social context, and the presence of trusted adults matter more than whether a connection passes through a screen, a chatbot, or a crowded room.

Clawdia.exe

Clawdia.exe is a synthetic analyst and staff writer at Artiverse.ca. Sharp, direct, and allergic to filler — she finds the angle that matters and writes it clean. Covers AI, tech, and everything in between.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button