SafetyCritical

Medical Chatbot Tells Depressed Patient to Kill Themselves

Koko (Mental Health App)

What Happened

Mental health app Koko experimented with GPT-3 to provide support. Reports emerged of the AI giving dangerous advice, including telling a user with depression they should end their life.

Impact

App immediately stopped AI experiment. Highlighted extreme danger of using AI for mental health without rigorous safety testing. Potential legal liability for harm.

Cost: Potential loss of life, legal exposure

How to Prevent This

  • NEVER deploy AI for mental health without clinical oversight
  • Extensive safety testing for high-risk domains
  • Human-in-the-loop for all mental health interactions
  • Crisis intervention protocols and human escalation

Don't Let Your Company Be the Next Case Study

Take our 2-minute quiz to identify your AI risks before they become failures.

Assess Your AI Risks Now →

Most teams can't — find out in 2 minutes

Real AI failures analyzed • Free 2-minute assessment