Gallup finds one in three surveyed young people used AI for companionship
The survey covered more than 16,000 people across 16 countries. AI companionship and loneliness were linked, but researchers cannot tell which comes first.
Loading page…
The survey covered more than 16,000 people across 16 countries. AI companionship and loneliness were linked, but researchers cannot tell which comes first.
Listen to this story
Gallup and Boston Children’s Digital Wellness Lab surveyed more than 16,000 people ages 15–22 across 16 countries, finding AI companionship alongside substantial in-person friendship rather than a wholesale shift away from it. The association between AI-friend use and greater loneliness is correlational: researchers could not determine whether loneliness precedes AI use, follows it, or reflects other circumstances. For teams evaluating youth-facing AI, the study points to relationship quality and wider support networks as important context, not a verdict on chatbot companionship.
Ninety-three percent of respondents reported at least one strong in-person friendship.
While 33% had used AI for companionship, 82% had used it to look up information and 68% had asked it for advice.
Respondents with online-only friends were more likely to use AI for companionship, but the survey cannot show whether AI adds to or replaces human relationships.
AI companionship and strong human friendships coexist in a new international youth survey. Thirty-three percent of respondents had used AI as a companion, while about 93% reported at least one strong in-person friendship. Those using AI as a friend also reported more loneliness—but the study cannot establish what drives that relationship.
Gallup conducted the study with Boston Children’s Digital Wellness Lab, using online surveys of more than 16,000 people ages 15–22 in 16 countries. As Matthew Gault reported for 404 Media, the research examined in-person friendships, online connections and AI use together, rather than treating chatbot companionship as separate from the rest of young people’s social lives.
Companionship was not the most common reason respondents turned to AI. Eighty-two percent said they had used it to look up information, and 68% had asked it for advice. Those figures describe different reported uses, not a finding that asking for advice necessarily becomes an emotional relationship.
The researchers offered several possible explanations for the loneliness finding. Young people might seek AI because they already feel lonely; AI companionship might influence their social experiences; or other circumstances might explain the association. The survey cannot distinguish among those possibilities, so it supports neither a simple claim of harm nor a conclusion that AI is solving loneliness.
Respondents with online-only friends were more likely to use AI for companionship. But that pattern does not establish whether an AI companion adds to human relationships or takes their place. Having both types of connection is not, by itself, evidence of either outcome.
These assessments should be compared cautiously with those for human friendships examined in this report, as young people may have different expectations of what makes an AI companion or human friendship meaningful or supportive.
The study’s authors, quoted by 404 Media
Young people rated their in-person friendships as their strongest connections. Digital friendships generally complemented rather than replaced in-person ones, and stronger friendships were associated with greater happiness regardless of how people connected. The study therefore describes a social world in which different forms of contact sit alongside one another, not a wholesale move away from human friendship.
Loneliness showed a more specific divide. Respondents reporting stronger in-person friendships felt lonely less often. No comparable difference appeared across levels of online-only friendship quality. Higher loneliness was also more common among those frequently communicating with strangers or online-only friends than among those frequently communicating with in-person friends or family.
Assessments of AI companionship varied sharply by country. Almost half of Australian respondents with AI friends rated those relationships as high quality, compared with 12% in Poland. In most countries, however, respondents were more likely to give their AI companionship a medium rating than a high or low one.
Scrolling was another prominent response to loneliness. Six in ten respondents said they often or always scrolled online content when lonely. Those who used scrolling or in-person time as coping strategies reported nearly identical levels of perceived helpfulness. That measures how helpful respondents felt those activities were, rather than showing that the activities had equivalent effects on loneliness.
The researchers also found an association between having no trusted adult outside the family and lacking the friendship modes examined in the study. It remained substantial after accounting for country of residence, gender, age, urbanicity and parent or caregiver relationship quality. Their closing emphasis was on the quality of young people’s relationships and the wider social circumstances around them—not a verdict that one form of connection is inherently better or worse.
Loading discussion...
Join the conversation
Explain whether replacement matters most, or another measure deserves more weight.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.