The marriage between AI and new datasets promised early detection of mental health problems and symptom prediction. This may enable timely and proactive instead of (traditionally) reactive interventions to mitigate mental health crises which could truly help both patients and healthcare systems. Yet, although the traditional healthcare model has been suboptimal to address the rising mental health problems, we are still not witnessing a long-awaited shift in the paradigm – from reactive care to preventative care. Are the algorithms ready? Are the healthcare systems ready? Is the world ready? This talk will delve into these questions by drawing from years of research at Koa Health and from a recent study that evaluated AI-based prediction of mental health crises.