Using AI to Break the Silence on Mental Health
“I grew up with the belief that admitting mental health issues is admitting you are weak. You must never show you are weak cause you lose face and people will look down on you,” said a Reddit user.
According to the WHO, nearly one billion individuals worldwide suffered from a mental illness in 2022. These alarming numbers are reflected online — with the search volume around ‘depression’ and ‘anxiety’ increasing by 35% in the last five years. However, the stigma around mental health is a major barrier to seeking support for those in need.
Social media is a game changer in breaking social barriers. Its communal and visual platforms have helped people voice their concerns, relate to others, and seek trusted resources. This also makes social media a powerful tool for governments and NGOs to reach out to those in need. However, the internet can be chaotic, and one can easily get lost looking for accurate patterns and insights. The process of analyzing social media data without proper tools can be expensive and time-consuming.
This is where AI can help. AI algorithms can analyze millions of data points quickly and efficiently to identify patterns and trends. Researchers can gain a deeper understanding of complex problems, identify new opportunities, and make more informed decisions using AI.
The SOAS Centre for AI Futures, in partnership with SOAS, Quilt.AI, and the Helsinki Institute of Social Sciences and Humanities, has developed AI-powered research capabilities to de-noise millions of Internet data points into insights for social good. We used our Culture AI to identify the dominant reasons behind mental health stigma using social media data.
We studied hundreds of social media posts about mental illness stigma from Reddit and TikTok to identify dominant themes and behaviors from across the world. The model categorized the posts into three domains based on the language and the content of the posts:
- Self: the negative attitudes, including internalized shame, that people with mental illness have about their own condition 
- Community or public stigma: Stigma created by the family, peers, and surrounding community 
- Professional: Stigma created at a workplace 
What did we observe?
Community-created stigma causes the most harm:
About 68% of the posts we studied talked about community-created stigma on TikTok. Individuals mostly flagged the jarring implications of the community-created stigma that prevented them from seeking support. Adolescents struggle the most from the mental health stigma imposed by their parents, while teens use social media platforms to seek advice on how to convey their mental health condition to their parents and seek professional support.
TikTok is a space to share and seek resources
Our model suggested that 78% of the posts talked about the existence of mental health stigma in society and the need to break the stigma, while only 17% of the posts had actionable insights on breaking the stigma in daily life.
TikTok is dominated by awareness-creating videos coupled with stories of personal triumph that are received with appreciation and compassion in the comment section. Most TikTokers share encouraging messages, asking their viewers to accept their illnesses. Evidently, people are also looking for resources in the comment sections. For example, individuals are asking specific questions such as ‘Can I ask how you obtained therapy, was it NHS or private,’ ‘but how to do it?’ etc., seeking actionable support to break the stigma.
Asians prominently express their frustration around mental health stigma online.
We observed that most mental health stigma-related subreddits originated from Asian countries, including the Philippines, India, and South Korea. Out of 14 subreddits, eight subreddits exclusively hosted posts from Asian countries. Individuals share their experiences about the lack of accessibility and acceptance of professional help due to cultural prejudices. Interestingly, Reddit posts about the mental stigma were dominated by males, whereas TikTok posts were skewed toward females.
We believe the Internet is the world’s largest focus group. As people turn to the Internet as the first stop for community and resources, it becomes a crucial intervention point to reach people who face stigma or discrimination offline. By using emerging technologies and AI tools, we can identify patterns and behaviors at a scale that informs future interventions. To address social issues more efficiently, it is crucial to embrace these new platforms and technologies.
To read more about AI models solving social problems, visit the Launch of SOAS Centre for AI Futures (Can AI And Anthropology work together?) in partnership with SOAS, Quilt.Ai, and the Helsinki Institute of Social Sciences and Humanities.