This article is part of our monthly Mental Health & Climate Change series about how extreme weather and other climate impacts affect our psychological well-being—and ways we can cope.
Let me preface this by saying that I’m a bit nervous about what I’m about to write: that despite my many concerns about artificial intelligence (AI) and the data centers needed to power it, I believe AI could play an important role in addressing the mental health impacts of climate change.
I’m worried about the environmental and human costs of AI for so many reasons. But I can tell you what changed my mind about the potential of well-designed AI tools to support mental health in an increasingly unstable climate—a conversation with Dr. Adrienne Heinz and my personal experience using AI tools to help navigate a season of crisis in my own life.
Over the past several months, Moms’ spring policy intern Bobby Cook and I each connected with Dr. Heinz, who describes herself as an “accidental disaster mental health specialist.” She’s a Stanford researcher and expert at the intersection of climate change and mental health and the co-founder of The After Collective, a platform for AI-powered disaster support. Thanks to a combination of personal and professional experience, Dr. Heinz is all too familiar with the challenging aftermath of disasters like wildfires, hurricanes, and floods: navigating fragmented aid systems, insurance company woes, and endless administrative barriers, all while dealing with disaster-related trauma and grief.
The After Collective platform focuses on providing “24-7 stress management and logistical support for anyone impacted by a disaster.” Dr. Heinz emphasizes that the platform uses a coaching approach and is not intended to replace therapy. Instead, the tool is designed to provide evidence-based mental health support that can help survivors cope with post-disaster stress, with the ability to connect users with real humans for additional support if needed. It can also support survivors in navigating the complicated red tape that is a hallmark of disaster recovery—insurance paperwork, resource coordination, document replacement.
We know that the mental health toll of climate disasters is already immense. Living through a climate disaster is linked with higher rates of depression, anxiety, post-traumatic stress disorder, and increased suicide risk. A heating planet means stronger and more damaging weather disasters, more moments of mental health crisis, and more people who will need mental health support.
We also know that our current mental health system is overtaxed in so many ways: there’s a severe, nationwide shortage of mental health providers, particularly in rural areas. In Vermont, where I live, good luck finding an in-person mental health provider without a monthslong waitlist who accepts your insurance and new patients. Vermont’s recent flood events exacerbated mental health needs that the state was already struggling to meet. This is a story unfolding across the nation: increasing mental health needs as an overstretched system fails to meet them.
I’ve never lost property or a loved one in an extreme weather disaster. But I’ve thought of Dr. Heinz’s work often as I care for a child with PANS/PANDAS, a chronic neuroimmune illness that can feel like an extended disaster for a family. Health crises, like disasters, don’t wait for convenient timing: they happen whether you’re prepared to deal with them or not.
If you’d asked me a year ago whether I’d ever consider turning to AI for personal support, I would have told you to take the nearest chatbot and put it where the sun doesn’t shine. (Well, I might have been more diplomatic than that.) The past year of my family’s life, though, has been marked by firsthand experience of the mental health provider shortage in Vermont. We’ve also faced a complicated insurance battle similar to what many disaster survivors face, including exhaustion, endless paperwork, and frequent retelling of the same distressing events to multiple insurance company representatives—only to encounter repeated denials. When a friend suggested to me several months ago that I try an AI-powered tool built to fight repeated insurance denials for needed PANS/PANDAS care, I was desperate. And I was startled by how useful it was.
I’ve also turned to a chatbot for help in prioritizing the incredibly complicated menu of tasks involved in my child’s medical care: multiple providers at three different clinics, a small pharmacy’s worth of medications. When we had to rush to the ER earlier this month, I used the chatbot to help me quickly generate a list of my child’s prescriptions and suggest productive ways to talk to difficult providers. It … helped.
The After Collective is still in early stages of development and not yet available for widespread individual use, but after my own experiences with AI in recent months, I’m convinced that there are many people a platform like The After Collective could help. Among the most promising potential benefits include broadening access to mental health support in areas where there simply aren’t enough providers, or where financial barriers put care out of reach.
None of my experiences with AI over the last year have eased my concerns with the enormous environmental impact of data centers, nor my worries about what AI could mean for human relationships over time. Yet I also know what it’s like to be desperately trying to help a loved one having a mental health crisis in the middle of the night, and what it’s like to fight a brutal insurance battle when you least have the inner resources to do so. If carefully designed tools like The After Collective can reach people in moments and ways that our fragile mental health care system can’t, that’s valuable—and it may be especially valuable for communities with the fewest resources and the greatest exposure to disaster.
At the end of the day, everyone deserves support in moving through the inevitable disasters of their lives, however they unfold. Figuring out how to make that happen, without adding to the harms AI already causes, is work worth doing. This means building out AI in a sustainable way with regulations to protect our health.
Many thanks to Bobby Cook, who contributed to the background research for this article.
Tell Congress: Protect Families From Dangerous Data Center Air Pollution