Everyone is asking how artificial intelligence will change humanity.

I think we're asking the wrong question.

The question that matters is whether AI will strengthen our capacity to care for one another, or quietly replace the human relationships and associated skills that make us resilient and enable us to flourish.

We're in a narrow but consequential window. AI capabilities are accelerating and being deployed far faster than the governance systems and institutions designed to guide them. The decisions we make today about how AI is built, deployed, and governed won't just determine the future of technology. They'll shape the future of human wellbeing.

As Chief Impact Officer at Crisis Text Line, I spend my days thinking about both.

Mental health is one of the clearest tests of responsible AI because the stakes couldn't be higher. When someone reaches out in their darkest moment, they don't need a chatbot to simulate care for them. They need another human being who knows how to listen and actually cares.

That's why we've made a deliberate choice: AI should augment human connection, not replace it.

At Crisis Text Line, AI helps us identify urgency, improve volunteer training, and operate at a scale that would otherwise be impossible. But every conversation with someone who reaches out is with a real, trained volunteer Crisis Counselor, supported in real time by a staff of mental health professionals.

Technology makes the speed and scale of connection possible. People make the connection meaningful.

This distinction matters. And the people we serve tell us exactly why that matters.

One texter shared, "Every word I typed was heard in a way no AI therapist could understand." Another told us, "Thank you for reminding me good, caring-hearted people still exist." These aren't just expressions of gratitude. They're reminders that empathy isn't a feature to be automated. It's a human experience, one that must be nurtured to generate and regenerate genuine, reciprocal care between human beings.

Today, our AI-undergirded, volunteer-powered model delivering free, confidential support is accessible to approximately 12% of the world's population across roughly 25 countries, with an ambition to reach 50% over the next three to five years.

Together, with our global partners, we've supported more than 18 million crisis conversations and trained more than 130,000 volunteers.

In the United States last year alone, approximately 80 mental health professionals on staff, including social workers, psychologists, and therapists, supported more than 14,000 volunteers. Those volunteers logged onto our purpose-built platform in two-hour shifts, 24 hours a day, and together supported more than 1.5 million conversations with people experiencing emotional distress.

Those numbers are significant. But they're not what excites me most. What inspires me is what happens after volunteers log off.

Our research suggests that the skills volunteers develop, such as empathy, active listening, collaborative problem-solving, emotional regulation, and crisis de-escalation, become part of how people show up in their own lives. They use those same skills with their families, partners, colleagues, classmates, neighbors, friends, and even strangers. They recognize signs of distress earlier. They navigate conflict differently. They create spaces where people feel seen.

Conservatively, we estimate that for every crisis conversation supported through Crisis Text Line, those volunteers help support or prevent another moment of crisis in their everyday lives. That means roughly 3 million moments of crisis supported or prevented in the United States just last year.

Think about what that represents. Not AI replacing human care. AI enabling more human care.

That, to me, is the multiplier effect we should urgently pursue.

We need leaders everywhere to rapidly architect new systems, pathways, partnerships, and guardrails to ensure AI delivers human wellbeing at scale.

‍

About the Author:

Margaret Meagher is Chief Impact Officer at Crisis Text Line, a global non-profit supporting 18+ million text-based mental health & crisis conversations worldwide since 2013. She leads AI/ ML/ NLP innovations and solutions-oriented research science to analyze millions of mental health conversations, and related impacts of its volunteer crisis responder training on 130k+ volunteers and their communities.