Ten percent of Americans consider an AI a close friend
Tracker summary: Ten percent of Americans, mostly young people, consider an AI a close friend.
A related friendship measure has reached 10%, but wording and population differ from this prediction.
At a glance
- Assessment: Emerging
- Confidence in assessment: 80%
- Outcome: unresolved
- Timing: pending
- Evidence: proxy
- Predicted timing: July 2027
- Primary source: July 2027: The Cheap Remote Worker, archived PDF p. 16
What AI 2027 Predicted
Tracker summary: Ten percent of Americans, mostly young people, consider an AI a close friend. The source is July 2027: The Cheap Remote Worker in the AI 2027 scenario.
Archived scenario version and section-page references. The linked live scenario may change; these definitions use the archived version identified here.
Scenario source locator: July 2027: The Cheap Remote Worker: archived PDF, p. 16.
How We Track This
Primary operational population: US adults, explicitly labeled as a limitation relative to Americans. Require a representative weighted survey of at least 1,000 adults using the exact close-friend concept, fielded by July 31, 2027, and independent corroboration with comparable wording within 60 days of the qualifying survey’s fieldwork end. A qualifying survey before July 2027 may resolve the test early once corroborated; a July survey is within-window. July 2027 remains the deadline for the threshold observation; corroboration may arrive afterward within the stated 60-day interval. The tracker uses 60 days to test reproducibility while limiting drift in a rapidly changing adoption measure; this is not a scenario requirement. Threshold: weighted point estimate at least 10%; publish uncertainty and age breakdown. Ask directly whether the respondent considers any AI a close friend.
These operational thresholds specify this tracker’s test. Where the scenario uses qualitative language, they are interpretations rather than numbers quoted from the original.
Current Evidence
YouGov’s July 1, 2026 report found that 10% of surveyed US adult citizens described an ongoing personal friendship with an AI chatbot, including 18% of adults under 30. Fieldwork was June 26-29, 2026, with 1,126 respondents recruited from YouGov’s opt-in panel and weighted by gender, age, race, education, 2024 presidential vote, 2020 election turnout and presidential vote, baseline party identification and voter registration. Weighting does not turn this into a probability sample, so the representative-survey requirement needs a sampling-quality assessment rather than an automatic pass. The age result indicates higher reported friendship among younger adults; it does not establish that most people with AI friends are young. This is useful precursor evidence, not the close-friend question. Its earlier date does not itself disqualify it.
Counterevidence & Limitations
Lower related estimates are counterevidence to treating the YouGov result as an established close-friend threshold, not directly comparable failures of the test:
- Elon University / YouGov, fielded May 18-22, 2026; university release September 2, 2026: the release reports social/emotional AI use among 27% of internet-using US adults. In that user subsample, 31% answered yes to “Do you consider AI to be a friend?” (topline, question 66, printed p. 85 / PDF p. 92). Multiplying the two rounded shares implies about 8.4% of internet-using adults (tracker calculation, not a published population estimate). The subsample is not all Americans, and “friend” does not mean “close friend.” The methodology and cover specify May 18-22; table footers inconsistently say May 18-23.
- Pew Research Center, fielded February 17-23, 2026; published June 17, 2026: the survey reports 4% of US adults ever using chatbots for companionship. The questionnaire, p. 14 asks “Do you ever use artificial intelligence (AI) chatbots for the following?”, including “For companionship”. This measures a use purpose rather than perceived friendship, so it cannot be converted into a close-friend prevalence or used as an independent replication of the YouGov question.
Adult-only measurement cannot directly establish the all-ages proportion. Chatbot usage, emotional attachment or companionship are not synonymous with close friendship. Report sampling and nonresponse uncertainty; do not infer causality or wellbeing effects.
Missing comparable survey evidence leaves the result unresolved.
What Would Change Our Assessment
Representative comparable surveys remain below 10%, or the result depends on leading wording or a selected AI-user sample.
Confirm → Confirmed: Evidence satisfying the full test resolves the threshold; an observation before July 2027 is scored as early timing once independently corroborated. An observation in July 2027 is scored within-window. Partial proxies are recorded with their scope.
Update History
| Date | Update |
|---|---|
| 2026-11-07 | Assessment reviewed. Status: Emerging; confidence: 80%. YouGov’s June survey found 10% reporting ongoing AI friendship and 18% among under-30s, supporting meaningful precursor adoption. Ongoing friendship is not the exact close-friend question; the opt-in sample and Pew’s lower companionship measure prevent treating this as confirmation of the population, age-composition and corroboration tests. Clarified survey populations and question wording, including YouGov’s under-30 prevalence rather than the age composition of all AI friendships. Added the lower Elon and Pew measures with their comparability and date caveats. Qualifying surveys may resolve before July 2027; the 60-day independent-corroboration rule remains explicit. The header is a scenario summary, not a quotation. Corrected the weighting-variable list to match the survey’s published methodology. |
| 2026-09-06 | A related friendship measure has reached 10%, but wording, population and observation window differ from this prediction. Relevant primary evidence and limitations are described above. |