Searchlight – Using Search Engine Data for Detection and Early Intervention in Suicide Prevention

At a glance

Funded by: NIMH 2021-2024
Principal Investigator(s): Kate Comtois Trevor Cohen
Research Team:
Participating Agencies:
Research Setting: Fully Online


Unlike social media which is curated for public view, what we search for on Google and YouTube is private, often with questions we are too shy to ask our friends or our doctor. Therefore, online search-engine behavior (Google and YouTube) may be an effective, private, and immediate method of suicide risk detection for anyone, regardless of their contact with systems of care. In this study, participants who made a suicide attempt in the past year, those who have made an attempt more than a year ago, and those who have thoughts of suicide but never attempted complete gold-standard research assessments of their suicidal thoughts and behavior and provide their Google Search and YouTube data via the publicly available Takeout function. We will identify and evaluate proximal risk factors in search-engine behavior; that is, patterns that change at the time of suicidal behavior or high suicidal ideation. We will also examine the ethical, legal and social implications of machine learning risk detection and the preferences of those with lived experience of suicidality for treatment and interventions when risk is detected.


  • We developed a web app to facilitate downloading Search and YouTube data from Google’s publicly available Takeout function.
  • A pilot study evaluated natural language processing as well as the timing and frequency of Google Search behavior relative to suicide attempts. 
  • Recruitment in progress

Learn More at The Searchlight Study and our ethics focused supplement study, Searchlight Ethics