"Dr. TikTok": How algorithms suggest ADHD and autism diagnoses

The internet allows patients to take a more active role in the diagnostic process, as the interactivity of "Dr. Google" and other online symptom-checkers demonstrates. Social media has emerged as a locus where (self-)diagnoses can be made through social interaction with others who identify as holding diagnostic identities. This paper builds on a foundation of research analyzing TikToks about ADHD and autism and the platform's content recommendation algorithm to examine how users are served content that implies that they occupy a given diagnosis, representing a distinct iteration of Dr. Google. The algorithm, not the user, could trigger the diagnostic process by labeling the user with a diagnosis or other pathologizing characteristics. I analyze how the TikTok algorithm encourages content themes and forms that may alter public perception of the boundaries, characteristics, and prevalence of ADHD and autism, as well as their entanglement with each other and the label of neurodivergence. I engage with the TikTok algorithm using digital ethnographic and discourse analysis approaches to understand "Dr. TikTok" as an example of how diagnostic pathways may be becoming increasingly mediated by AI systems. I find that the TikTok algorithm shapes TikToks about ADHD and autism by encouraging creators to engage in practices like speaking in second-person to the viewer and positioning relatable experiences as a form of subjective diagnostic expertise, situating these videos as paradoxically hyperpersonalized, yet inclusive. This study contributes to the sociology of medicine by showing how AI systems could intervene in and reorder the traditional diagnostic process.