Typology Of AI-Based Therapy Micro-Bursts Reveals How People Lean Into Chatbots For Mental Health Advice
I've previously defined AI-based therapy micro-bursts, which is how people tend to use AI for mental health advice. Now, here's my typology. An AI Insider scoop.
- Lance Eliot’s typology identifies five AI therapy micro-burst patterns: crisis check-ins, venting sessions, reflective prompts, advice-seeking, and goal-tracking.
- More than 40 million people downloaded mental health chatbots in 2025, a 60% increase from 2024 (Sensor Tower).
- 70% of AI therapy micro-bursts last under 5 minutes, with 45% occurring outside standard clinic hours (between 10 p.m. and 6 a.m.).
- Generative AI models like GPT-4 and Gemini now power 85% of therapy chatbots, up from 30% in 2023 (Grand View Research).
- A 2025 Stanford study found that structured micro-burst interactions reduced self-reported stress scores by 18% within two weeks of use.
Frequently Asked Questions
AI therapy micro-bursts are short, fragmented interactions where people use generative AI chatbots for on-demand mental health advice. They typically last under five minutes and serve purposes like crisis check-ins, venting, or reflective prompts.
Lance Eliot, a Forbes contributor and AI ethicist, defined and categorized AI therapy micro-bursts into five patterns: crisis check-ins, venting sessions, reflective prompts, advice-seeking, and goal-tracking.
Over 40 million people downloaded mental health chatbot apps in 2025, a 60% increase from 2024. More than 80% of such chatbots now use large language models like GPT-4 or Gemini.
Early studies show that structured micro-bursts can reduce stress by about 18% over two weeks. However, experts caution about over-reliance, privacy risks, and lack of regulation, calling for more clinical trials.
Crisis check-ins handle urgent emotional peaks; venting sessions allow users to release frustration; reflective prompts encourage self-examination; advice-seeking targets specific problems; goal-tracking monitors progress. Each has distinct user triggers and bot responses.
Topics
Original source
www.forbes.com
Discussion
Join the discussion
Sign in to post a comment or reply.
No comments yet. Be the first to share your thoughts!