The Line Between Helpful and Harmful: How Responsible AI looks like in Mental Health Coaching with Shelby Garay

As Program Director of the Headspace Training Institute — training the next generation of mental health coaches — Shelby sits at one of the most critical intersections in coaching today: where AI meets mental health, and where getting it wrong has real consequences. This conversation is about what getting it right actually looks like. At Headspace, that means clinical safety guardrails with human oversight, a 30-minute limit on AI companion conversations to prevent overuse and prompt degradation, and an empathetic AI companion called Ebb that meets people where they are — every single day, not just in crisis. Because that's the shift Shelby keeps coming back to: mental health care can't only happen when things fall apart. It needs to be daily. Like going to the gym. Rebecca and Shelby also dig into how Headspace is training coaches to have the AI conversation with their clients — not avoid it — and why the coaching industry needs to stop asking "AI or human?" and start asking how AI can amplify what only humans can do. And in a moment that lands quietly but hard: Shelby shares why she's teaching her students to embrace failure — and why that might be the most human skill of all. In this episode: Why Headspace put a 30-minute limit on its AI companion — and why that's revolutionary How to build clinical guardrails that actually hold Training coaches to talk to clients about AI — not around it The danger of general-purpose chatbots being used for mental health Embracing failure as the foundation of mental health and coaching Links and resources mentioned in this episode: Headspace Training Institute Headspace Ebb - Headspace's Empathetic AI Companion Headspace Coaching Shelby Garay LinkedIn Shelby Garay Coaching & Consulting Find out more: https://vivalacoaching.com/vivapodcast/