You are using an outdated browser. Please upgrade your browser to improve your experience and security.

Skip to main

AI Chatbot Use for Mental Health Among LGBTQ+ Young People

Across two survey samples, LGBTQ+ young people reported using AI tools and applications, including for mental health support. In The Trevor Project’s 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People, 76% reported having ever used AI tools and applications and in Project SPARK, 40% reported having ever used an AI chatbot for mental health support.
Donate

Key Findings

  • In The Trevor Project’s 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People, most (76%) reported having ever used AI tools or applications, and a larger share (78%) reported being at least somewhat uncomfortable with the increasing integration of AI in daily life. 
  • In the Trevor Project’s longitudinal study, Project SPARK, AI use for mental health support varied by gender identity with cisgender participants reporting higher use (59%) than their transgender, nonbinary, or gender questioning (TGNB) peers (19%). 
  • Among Project SPARK participants who used AI for mental health support, most reported finding it somewhat (53%) or very (20%) helpful.

Background

The use of artificial intelligence (AI), including generative AI tools such as interactive AI chatbots that can simulate human conversations in real time, is becoming increasingly popular. Approximately two-thirds of youth aged 13 to 17 years in the United States (U.S.) have ever used an AI chatbot1 and nearly one in five U.S. youth and young adults aged 12 to 21 years have used an AI chatbot for mental health advice.2

AI chatbots may have potential benefits and risks regarding mental health support. One main advantage is that they may mitigate barriers to care as they provide easy and convenient access, are often available at little to no cost, and allow users to seek information about sensitive topics such as mental health concerns without needing to interact with a counselor or mental health care professional.3 AI chatbots also offer a form of social support and emotional validation, which may be beneficial for addressing mental health concerns.4 Finally, for people with multiple marginalized identities who often encounter multiple forms of bias and stigma when accessing mental health care,5 interacting with AI chatbots may reduce these adverse experiences.

There are several concerns about AI chatbots, mainly with regards to the inclination for and accuracy of the information they provide. Most AI chatbots were developed, trained, and tested using data and algorithms that were not representative of all people and experiences.6,7 For instance, AI chatbots developed without input from young people may provide information that is less relevant or appropriate for them than for adults.8 Further, training data for many general-purpose AI chatbots are not clinically validated and, as a result, AI chatbots may provide inaccurate and non evidence-based mental health advice.9 Finally, use of AI chatbots for mental health may increase dependency on them particularly because chatbots are trained to be agreeable and validate a user’s experience,9 which may perpetuate use of chatbots while also potentially discouraging some users from seeking help from a trained counselor or mental health professional.

Research on the use of AI chatbots for mental health is emerging, however, little is known about AI use patterns among LGBTQ+ young people. This may be particularly important considering LGBTQ+ people experience more mental health concerns than their non-LGBTQ+ peers,10 face a myriad of barriers when accessing mental health care,11,12 and often turn to virtual spaces to find social connection and support.13,14 These challenges are further exacerbated by a lack of available affirming care given limited training in LGBTQ+ health among mental health professionals.15

Using data from The Trevor Project’s 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People (N = 16,667) and Project SPARK: A Longitudinal Study of Protective and RisK Factors in LGBTQ+ Youth Mental Health (N = 788), this brief: 

  1. 1) examines demographic characteristics associated with AI use;
  2. 2) describes experiences with AI use for mental health support;
  3. 3) evaluates the relationship between access to mental health care and AI use for mental health.

Results

Results from the 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People

Demographic Differences in Use of AI Tools and Applications
In the 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People, 76% of participants (n = 11,227) reported having ever used AI tools or applications for any purpose.

  • By age: 13 to 17-year-olds were more likely to report AI use than 18 to 24-year-olds (84% vs. 72%). 
  • By gender identity: Cisgender participants were more likely to report ever using AI tools and applications (85%) than their transgender, nonbinary, or gender questioning (TGNB) peers (72%).


Experiences with AI Tools and Applications
The three most commonly used AI tools and applications were:

  1. 1) Chatbots or AI chat assistants (e.g., ChatGPT, Replika) (54%);
  2. 2) AI-generated answers in search engines (e.g., Google’s AI summaries, Bing AI) (42%);
  3. 3) AI recommendations for videos or music (e.g., YouTube “Up Next”, Spotify playlists) (36%).
Use of AI Tools and Applications chart

Despite its widespread use, most LGBTQ+ young people reported being uncomfortable with the increasing integration of AI in daily life, with 54% reporting they were “very uncomfortable” and another 24% reporting they were “somewhat uncomfortable”.

Comfort with Increasing AI Integration in Daily Life chart



Results from Project SPARK: A Longitudinal Study of Project and RisK Factors in LGBTQ+ Youth Mental Health

Demographic Differences in AI Use for Mental Health Support
In Project SPARK, 40% of LGBTQ+ young people (n = 316) reported having ever used an AI chatbot for mental health support. Participant demographic characteristics were associated with having ever used AI for mental health support:

  • By age: 18 to 24-year-olds were more likely to report use than 13 to 17-year-olds (42% vs. 26%). 
  • By sexual orientation: Gay (68%) and lesbian (56%) young people were the most likely to report having ever used AI for mental health support.
  • By gender identity: Cisgender participants reported having ever used AI for mental health support more compared to fewer than one in five of their TGNB peers (59% vs. 19%, respectively).
  • By race/ethnicity: Black/African American (81%) and Hispanic/Latinx (61%) LGBTQ+ young people reported the highest use of AI for mental health support than other groups (see Data Table).


Access to Mental Health Care and AI Use for Mental Health.
Compared to those who did not use AI for mental health support, participants who used AI for mental health support were less likely to have received helpful psychological or emotional counseling from a counselor or mental health care professional in the past 6 months (68% vs. 32%, respectively). Among LGBTQ+ young people who did not access mental health in the past 6 months, those who used AI for mental health support had fewer barriers to accessing mental health care (e.g., “I did not feel like they would understand my sexual orientation or gender identity” and “I did not want to have to get my parent’s/caregiver’s permission”) than those who did not use AI for mental health support (Mean [M] and Standard Deviation [SD]; M=0.6, SD ± 1.5 vs. M=1.4, SD ± 2.5, respectively).

Experiences with AI Use for Mental Health
Among LGBTQ+ young people who reported ever using AI for mental health support (n = 316), most reported their experiences as helpful, with 53% reporting it was “somewhat helpful” and another 20% reporting it was “very helpful”. Further, while 29% of all LGBTQ+ young people surveyed reported that AI provided them with information to access a crisis or support hotline, this nearly doubled (60%) among those LGBTQ+ young people who had ever used AI for mental health support. Of those who used AI for mental health support and were provided with this information (n=187), 33% accessed the resource provided.

Experiences with AI for Mental Health Support chart

Looking Ahead

This research brief examined use of AI tools and applications for general purposes and use of AI for mental health support among LGBTQ+ young people. Existing evidence on AI use among youth and young adults is limited but emerging, however, evidence on patterns of use among LGBTQ+ young people is largely absent. These findings add to this growing research. Information about AI use for mental health is vital for supporting the needs of LGBTQ+ young people, considering they often experience a high burden of mental health concerns and barriers to mental health care.

Among those who participated in The Trevor Project’s 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People, we found that two-thirds (76%) had ever used AI for any purpose, with more than half of participants (54%) using AI chatbots – the most commonly used AI tool or application. In this sample, use of AI varied by age and gender identity with self-reported use being highest among youth (13 to 17-year-olds) and cisgender participants. Although use of AI tools and applications was common, more than two-thirds of participants reported being either somewhat or very uncomfortable with the increasing integration of AI in daily life.

We also found differences in AI use for mental health support using data from The Trevor Project’s longitudinal study, Project SPARK, with 40% of LGBTQ+ young people reporting having ever used AI for their mental health. Young adults (18 to 24-year-olds), cisgender, gay and lesbian, and Black/African American and Hispanic/Latinx participants were more likely to use AI for mental health support. Among LGBTQ+ young people who had used AI for mental health support, nearly three quarters reported their experiences as at least somewhat helpful and 60% were provided with information to access a crisis or support hotline resource. This may suggest a role for AI chatbots in facilitating connections to these resources, however, future research should explore whether these resources are affirming and appropriate for LGBTQ+ young people. Additionally, more research is needed to understand how use of AI chatbots for mental health support may influence seeking help from a counselor or mental health care professional. We found that LGBTQ+ young people who used AI for mental health support were less likely to receive counseling in the past 6 months but also reported fewer barriers to accessing mental health care. Future research that elucidates whether AI chatbots are supplementing or serving as an alternative to professional mental health care is needed.

Our findings align with the limited evidence on AI use among youth and young adults. Prior research has found that AI use for mental health was highest among 18 to 21-year-olds compared to 12 to 14-year-olds,2 which aligns with our findings indicating older participants were more likely to use AI for mental health support while younger participants were more likely to use AI for general purposes. We also found that AI use was more common among cisgender participants than their TGNB peers. Prior evidence suggests that transgender young people are less likely to use AI tools or applications, with many expressing concerns about AI-perpetuated harassment and bias.13 Finally, AI chatbots for general purposes and for mental health support have also been found to be used more among Black/African American, Hispanic/Latinx, and youth who listed a race/ethnicity other than White similar to our findings.1,2,13

Future research should continue to explore patterns of AI use among LGBTQ+ people of color to understand how to better support their mental health needs, particularly if they are turning to virtual modalities to address these needs. In particular, qualitative research studies are warranted. For example, such studies can provide a richer understanding of the reasons why LGBTQ+ people of color are using AI for mental health more than their peers and also provide an understanding of their experiences using AI. It is possible that AI use mitigates the historical barriers LGBTQ+ people of color have faced when accessing mental health care,5 thereby reducing inequities in accessing support. However, LGBTQ+ people of color may be at increased risk for overreliance on AI chatbots if they are the main source of mental health support, which may perpetuate inequities.9 

There are a few limitations to this research. Although the sample from the 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People is large and diverse, it is not intended to be representative of all LGBTQ+ young people in the U.S. and additional research is needed at the population-level to understand AI usage among youth and young adults. Second, to describe AI use for mental health, this research brief only used data from Wave 6 of Project SPARK, which was the first data collection wave that included these measures. Future data collection waves that assess AI use for mental health will allow us to examine trends in use over time. Despite these limitations, this research brief may be one of the few to present findings on AI use for mental health support among LGBTQ+ young people.

The Trevor Project’s free, crisis services are available 24/7 and are committed to ensuring that all LGBTQ+ young people receive high-quality care that addresses the unique experiences they face, including mental health concerns. The Trevor Project trains crisis counselors to provide culturally competent and non-judgemental support. The TrevorSpace platform is a safe and supportive online environment for LGBTQ+ young people ages 13 to 24, where they can connect with peers, find support and community, and share their experiences related to mental health.

To read more on this topic, check out our brief on Online Experiences and Mental Health of LGBTQ+ Young People. Additionally, to read more about online safety to help navigate the internet more securely including when seeking supportive communities, check out our resource guide.

Data Tables

Participant Characteristics of LGBTQ+ Young People in The Trevor Project’s Project SPARK Wave 6, by use of AI for Mental Health

Participant Characteristics of LGBTQ+ Young People in The Trevor Project’s Project SPARK Wave 6, by use of AI for Mental Health table

Methods

Sample
Data were collected through The Trevor Project’s 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People and Project SPARK: A Longitudinal Study of Protective and RisK Factors in LGBTQ+ Youth Mental Health. For the U.S. National Survey, 16,667 LGBTQ+ young people between the ages of 13 to 24 were recruited via targeted social media ads, posts by LGBTQ+ “influencers”, state and national organizations that serve LGBTQ+ young people, and direct outreach efforts (e.g., The Trevor Project’s e-mail listservs). For Project SPARK, 1,689 LGBTQ+ young people between the ages of 13 to 24 were recruited via social media ads to participate in a baseline survey (Wave 1) and follow up surveys every six months (Waves 2-6). This research brief used data from Wave 6 of Project SPARK, which included 788 LGBTQ+ young people.

Measures
Demographic characteristics (i.e., age, sexual orientation, gender identity, race/ethnicity) were assessed by asking participants to select a response from a provided list of options.16,17

AI use was assessed by asking participants in the 2025 U.S. National Survey, “How often do you interact with AI-powered tools or applications” and participants in Project SPARK, “How often have you used AI or a chatbot for mental health support?”. For both questions, response options included: “Never”, “Rarely (less than once a month)”, “Occasionally (1-3 times a month)”, “Frequently (1-3 times a week)”, and “Daily”. For this brief, responses were categorized as either ever (less than once a month to daily) or never.

Participants in the 2025 U.S. National Survey were also asked “Which of the following AI (artificial intelligence) tools or applications have you used? Select all that apply.” Response options include 9 types of tools or applications (e.g., AI chatbots or writing helpers). Finally, participants were asked, “How comfortable do you feel about the increasing integration of AI in daily life?”. Response options included: “very uncomfortable”, “somewhat uncomfortable”, “neutral”, “somewhat comfortable”, and “very comfortable.”

Participants in Project SPARK were asked about their experiences with AI for mental health using the following question, “Overall, how helpful have your experiences with AI or chatbots been when you needed mental health support?”. Response options included: “Very unhelpful”, “Somewhat unhelpful”, “Neither helpful nor unhelpful”, “Somewhat helpful”, and “Very helpful”. Additionally, participants were asked, “Has AI or a chatbot ever provided you with information to access a crisis or support hotline?” with responses options “No”, “Yes, but I have never accessed a crisis or support hotline this way”, and “Yes, and I have accessed a crisis or support hotline this way.”

In Project SPARK, participants were also asked about access to mental health care using questions including, “In the past 6 months, have you ever received psychological or emotional counseling from a counselor or mental health care professional?” with response options: “No, I do not want it”, “No, but I would like to receive it”, “Yes, but it isn’t helpful”, and “Yes, and it is helpful”. For this brief, analyses were limited to those who indicated “Yes, and it is helpful”. Participants were also asked, “Did you not see a counselor or mental health professional for any of the following reasons in the past 6 months? Please select all that apply.” Response options included 23 reasons (e.g., “I could not afford it” or “They were too focused on my LGBTQ+ identity”). For this brief, a total sum of reasons was calculated to assess the number of barriers to accessing mental health care.

Analysis
Chi-squared tests were run to examine differences between groups. Unless otherwise noted, all reported differences were statistically significant at p<0.05, meaning the probability of these differences being by random chance would be expected less than 5% of the time even if there were no true differences (i.e., under the null hypothesis). Percentages in the data table may not add up to 100 because of rounding and/or missingness.

References

  1. 1. Pew Research Center (2025). Teens, Social Media and AI Chatbots 2025. Washington, District of Columbia. https://www.pewresearch.org/internet/2025/12/09/teens-social-media-and-ai-chatbots-2025/#use-of-ai-chatbots

    2. McBain, R.K., Cantor, J.H., Breslau, J., Diliberti, M., Zhang, L.A., Zhang, F., Burnett, A., Kofner, A., Rader, B., Pataranutaporn, P., Stein, B.D., Mehrotra, A., & Yu, H. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics. 180(8):884-890. https://doi.org/10.1001/jamapediatrics.2026.2015

    3. Bragazzi, N. L., Crapanzano, A., Converti, M., Zerbetto, R., & Khamisy-Farah, R. (2023). The Impact of Generative Conversational Artificial Intelligence on the Lesbian, Gay, Bisexual, Transgender, and Queer Community: Scoping Review. Journal of Medical Internet Research, 25, e52091. https://doi.org/10.2196/52091

    4. Merrill, K., Mikkilineni, S. D., & Dehnert, M. (2025). Artificial Intelligence Chatbots As A Source of Virtual Social Support: Implications For Loneliness and Anxiety Management. Annals of the New York Academy of Sciences. 1549(1), 148–159. https://doi.org/10.1111/nyas.15400

    5. Ito, S., Jans, L. K., Rosenfeld, E. A., Shen, J., Vilkin, E., Gonzalez, A., & Vivian, D. (2025). Clinical Psychology: Science and Practice. 32(2): 105–121. https://doi.org/10.1037/cps0000241

    6. Tay L., Woo S.E., Hickman L., Booth B.M., D’Mello S. (2022). A Conceptual Framework for Investigating and Mitigating Machine-Learning Measurement Bias (MLMB) in Psychological Assessment. Advances in Methods and Practices in Psychological Science. 5(1). https://doi.org/10.1177/25152459211061337

    7. Landers, R.N., & Behrend, T.S. (2023). Auditing the AI Auditors: A Framework for Evaluating Fairness and Bias in High Stakes AI Predictive Models. American Psychologist. 78(1): 36-49. https://doi.org/10.1037/amp0000972

    8. American Psychological Association. (2025a). Artificial Intelligence and Adolescent Well-Being. Washington, District of Columbia. https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-ai-adolescent-well-being

    9. American Psychological Association. (2025b). Use of Generative AI Chatbots and Wellness Applications for Mental Health. Washington, District of Columbia. https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-chatbots-wellness-apps

    10. Wittgens, C., Fischer, M. M., Buspavanich, P., Theobald, S., Schweizer, K., & Trautmann, S. (2022). Mental Health in People with Minority Sexual Orientations: A Meta-Analysis of Population-Based Studies. Acta Psychiatrica Scandinavica, 145(4), 357–372. https://doi.org/10.1111/acps.13405

    11. Green, A.E., Price-Feeney, M. & Dorison, S. (2020). Breaking Barriers to Quality Mental Health Care for LGBTQ Youth. New York, New York: The Trevor Project. https://www.thetrevorproject.org/research-briefs/breaking-barriers-to-quality-mental-health-care-for-lgbtq-youth/

    12. The Trevor Project. (2024). Mental Health Care Access and Use among LGBTQ+ Young People. https://doi.org/10.70226/TERU5243

    13. Gardner, S. (2026). ctrl+alt+lgbt: Digital Access, Usage, and Experiences of the LGBTQ+ Community. Staunton, Virginia: LGBT Technology Institute (LGBT Tech). https://www.lgbttech.org/post/ctrl-alt-lgbt-new-national-poll-explores-lgbtq-digital-life-and-experiences 

    14. The Trevor Project (2025). Online Experiences and Mental Health of LGBTQ+ Young People. https://doi.org/10.70226/DZPX7792

    15. Jauregui, J.C., & Harper, G.W. (2025). LGBTQ+ Cultural Sensitivity Training for Mental Health Professionals in the USA. Archives of Sexual Behavior. 54(4): 1309-1315. https://doi.org/10.1007/s10508-025-03132-3

    16. Nath, R., Matthews, D.D., Hobaica, S., Eden, T.M., DeChants, J.P., Taylor, A.B., Suffredini, K. (2026). Project SPARK Two-Year Findings: A Longitudinal Study of Risk and Protective Factors in LGBTQ+ Youth Mental Health (2023-2026). West Hollywood, California: The Trevor Project. https://doi.org/10.70226/VFKG1732 

    17. Nath, R., Matthews, D.D., Hobaica, S., DeChants, J.P., Eden, T.M, Taylor, A.B., & Suffredini, K. (2026). 2025 U.S. National Survey on the Mental Health of LGBTQ+ Young People. West Hollywood, California: The Trevor Project. https://doi.org/10.70226/EKGT3197

The Trevor Project. (2026). AI Chatbot Use for Mental Health Support Among LGBTQ+ Young People. https://doi.org/10.70226/MZUJ5655

For more information please contact: [email protected]

© The Trevor Project 2026