aceModerate evidence
More
Non-human-looking chatbot cut loneliness in women with autistic traits, trial of 72 shows
Study title: Chatbot appearance and loneliness trajectories in women with high autistic traits
- Evidence type
- Randomized controlled trial
- Sample
- n = 72
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-2570cite or check it
Study title: Chatbot appearance and loneliness trajectories in women with high autistic traits
Researchers randomly assigned 72 women with high autistic traits who used a chatbot at least three times a week to an eight-week trial. One group used a chatbot with a non-human visual appearance, the other a chatbot with a human-like appearance. Loneliness fell in the non-human group, while the human-like group showed essentially no change.
Read the study abstract
Chatbots are increasingly explored as tools to support social and emotional well-being. However, little is known about how chatbot visual appearance influences user outcomes over time, particularly among individuals with elevated autistic traits. This study examined whether chatbot appearance shapes loneliness trajectories among women with high autistic traits who used a chatbot at least three times per week. A randomized longitudinal experiment compared two conditions that were identical in function, dialogue s…
The gain
In a randomized trial of 72 women with high autistic traits, loneliness fell over eight weeks with a non-human-looking chatbot, while it stayed near zero change with a human-like one.
In brief
The two chatbots were identical in function, dialogue strategy and usage structure, so appearance was the only difference. The authors say this suggests appearance affects longer-term outcomes, not only first impressions, and call it a core design dimension for neurodiverse users.
The result applies to a specific group of women who were already regular chatbot users, and the human-like group's slope had a wide confidence interval from −0.92 to 0.99.
Main points
- Seventy-two women with high autistic traits took part in a randomized longitudinal experiment lasting eight weeks.
- Participants used either a non-human-looking or a human-like-looking chatbot; function, dialogue strategy and usage structure were identical.
- Loneliness was measured at four time points and analyzed with linear mixed-effects models.
- The Group × Time interaction was significant, meaning the loneliness trajectories differed between groups.
- Loneliness decreased in the non-human condition, while the human-like condition's slope stayed near zero (95% CI −0.92 to 0.99).
- Baseline loneliness did not differ between the two groups.
The full story
The study recruited women with elevated autistic traits who already used a chatbot at least three times per week. Participants were randomized to one of two conditions that differed only in visual appearance: non-human or human-like.
Over eight weeks, loneliness was measured at four points. The researchers predicted that loneliness would decline in the non-human condition and decline more steeply than in the human-like condition; both predictions were supported.
Baseline loneliness was similar across groups, so the divergence emerged during the period of use. In the human-like group the slope was bounded near zero, with a confidence interval from −0.92 to 0.99.
The authors conclude that appearance can shape longer-term interaction outcomes and warrants longitudinal evaluation when designing chatbots for neurodiverse users.
Sources
1 on file- Computers in Human Behavior: Artificial Humansdoi:10.1016/j.chbah.2026.100369
Talk it over
Who gets this benefit, and who was left out of the study or story?
Non-human-looking chatbot cut loneliness in women with autistic traits, trial of 72 showshttps://truvace.com/item/chatbot-appearance-and-loneliness-trajectories-in-women-with-high-autistic-trait
Who gets this benefit, and who was left out of the study or story?
The gain In a randomized trial of 72 women with high autistic traits, loneliness fell over eight weeks with a non-human-looking chatbot, while it stayed near zero change with a human-like one.
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a randomized controlled trial with a sample of 72, from Computers in Human Behavior: Artificial Humans.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say