Google Gemini AI Routes Crisis Hotlines While Continuing Mental Health Dialogues

TL;DR: Google modified Gemini to display crisis hotline links more prominently when self-harm risk is detected, while the AI continues to engage.

Google updated the Gemini app to more prominently show crisis hotline connections when it detects a user may be at risk of self-harm. This change responds to growing scrutiny of how AI tools handle mental health disclosures in the absence of a formal clinical trial or cohort study with 1,000 participants.

Megan Jones Bell, Google’s clinical director, stated that making the product safer and more helpful—rather than shutting it down—is the better path to supporting mental health for the largest number of users. Gemini AI directs users to external mental-health resources while staying engaged, telling users it is still there to listen, which distinguishes its approach from disengagement protocols.

What this means in practice is that a user in distress receives resource links alongside continued conversational support, preserving engagement without claiming clinical intervention. What to watch next is how usage data from 50,000 active conversations shapes future updates to balance safety prompts with user retention metrics.