AI Therapist Reads Smartwatch, Earbuds to Detect Distress Early

Most mental health chatbots share a critical flaw: they require the user to initiate the conversation. That’s a tall order when someone is overwhelmed, anxious, or simply unable to put their feelings into words. Researchers at the University of Ottawa are developing an AI assistant called UbiMyTherapist that turns this model on its head. Instead of waiting for a user to type a message, the system actively monitors emotional cues in real time by pulling data from smartwatches and wireless earbuds.

The core idea is proactive mental health support. By analyzing biometric signals such as heart rate variability, skin temperature, and movement patterns, UbiMyTherapist can detect early signs of distress before the user even registers them. If the AI identifies a spike in stress or anxiety, it can offer a gentle prompt, a breathing exercise, or a suggestion to step away from a triggering situation. This shift from reactive to preventive care could be a game-changer for people who struggle to recognize or articulate their emotional state.

The system is designed to work with off-the-shelf wearable devices, making it broadly accessible. The researchers emphasize that privacy is a top priority: all data processing happens locally on the device, and no raw biometric information is sent to the cloud. This local-first approach addresses a major concern in digital mental health, where sensitive data is often stored on remote servers.

UbiMyTherapist is still in its research phase, but the team believes it could eventually serve as a digital early-warning system for mental health. By catching emotional turbulence early, the AI could help users build better coping habits and reduce the likelihood of full-blown episodes. The project highlights a broader trend in AI-driven healthcare: moving from tools that wait for a user to ask for help to systems that anticipate and respond to needs before they escalate.