Abstract
Background:
Social media platforms have become prominent spaces for health communication. Artificial intelligence (AI), particularly large language models (LLMs), is increasingly used to generate and mediate health-related content on these platforms. However, empirical evidence on how AI-mediated health communication functions in these environments remains limited and fragmented.
Objective:
This scoping review aimed to map empirical research on AI-mediated health communication on social media, focusing on communication modalities, evaluation dimensions (including quality, empathy, readability, and trust), and governance considerations.
Methods:
A scoping review was conducted in accordance with PRISMA-ScR guidelines. PubMed, Scopus, and Web of Science were searched for English-language empirical studies published between 1 January 2020 and 30 June 2026. Eligible studies examined AI-generated, AI-assisted, AI-analysed, or AI-related health communication within social media and online community environments. Data were synthesised using descriptive and thematic analysis.
Results:
Eleven studies were included, primarily examining communication in publicly accessible social media and online community environments. Studies were classified according to their primary communication function as response generation (n = 4), analytic interpretation (n = 6), or information structuring (n = 1). Evaluations focused mainly on empathy and humanistic care, readability, information and response quality, user perception and trust, bias and fairness, and the validity of AI-assisted analytic outputs. Considerable heterogeneity was observed in measurement approaches. Governance considerations were inconsistently operationalised and were generally secondary to the principal analyses, while accountability mechanisms and long-term relational outcomes remained largely unexamined.
Conclusions:
Existing evidence positions AI as a communicative intermediary that may have the potential to enhance accessibility and clarity in social media–based health communication. However, the literature is methodologically fragmented and offers limited insight into ethical governance and long-term implications. More standardised and longitudinal research is needed to support safe, accountable, and effective integration of AI in digital health communication.
Social media platforms have become prominent spaces for health communication. Artificial intelligence (AI), particularly large language models (LLMs), is increasingly used to generate and mediate health-related content on these platforms. However, empirical evidence on how AI-mediated health communication functions in these environments remains limited and fragmented.
Objective:
This scoping review aimed to map empirical research on AI-mediated health communication on social media, focusing on communication modalities, evaluation dimensions (including quality, empathy, readability, and trust), and governance considerations.
Methods:
A scoping review was conducted in accordance with PRISMA-ScR guidelines. PubMed, Scopus, and Web of Science were searched for English-language empirical studies published between 1 January 2020 and 30 June 2026. Eligible studies examined AI-generated, AI-assisted, AI-analysed, or AI-related health communication within social media and online community environments. Data were synthesised using descriptive and thematic analysis.
Results:
Eleven studies were included, primarily examining communication in publicly accessible social media and online community environments. Studies were classified according to their primary communication function as response generation (n = 4), analytic interpretation (n = 6), or information structuring (n = 1). Evaluations focused mainly on empathy and humanistic care, readability, information and response quality, user perception and trust, bias and fairness, and the validity of AI-assisted analytic outputs. Considerable heterogeneity was observed in measurement approaches. Governance considerations were inconsistently operationalised and were generally secondary to the principal analyses, while accountability mechanisms and long-term relational outcomes remained largely unexamined.
Conclusions:
Existing evidence positions AI as a communicative intermediary that may have the potential to enhance accessibility and clarity in social media–based health communication. However, the literature is methodologically fragmented and offers limited insight into ethical governance and long-term implications. More standardised and longitudinal research is needed to support safe, accountable, and effective integration of AI in digital health communication.
| Original language | English |
|---|---|
| Article number | 101293 |
| Pages (from-to) | 1-9 |
| Number of pages | 9 |
| Journal | Health Policy And Technology |
| Volume | 15 |
| Issue number | 9 |
| Early online date | 27 Jul 2026 |
| DOIs | |
| Publication status | Published - Nov 2026 |
Fingerprint
Dive into the research topics of 'Artificial Intelligence–Mediated Health Communication on Social Media: A Scoping Review'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver