AI is everywhere. From our inboxes to our investments, it’s the buzzword reshaping industries. But when it comes to healthcare and more specifically, housing stability for vulnerable populations AI’s promise isn’t just about automation. It’s about amplification: amplifying human connection, efficiency, and trust.
In our work with Medicaid members, older adults, and others navigating complex social needs, we’ve learned that engagement is everything. If we can’t reach someone, we can’t help them.
Historically, engagement meant downloading an app or navigating a portal. Barriers that often exclude the very people who need the most support. But AI changes that. We now use it to meet members where they are, through channels they already use, in ways that feel natural and human. It can adapt its tone, guide someone through a complex eligibility question, or simply help break the ice.
Used thoughtfully, AI becomes the “tip of the spear.” It doesn’t replace people, it just gets them ready to connect with the right one.
The second area where AI makes a big impact is operational efficiency. Housing services are full of tedious but essential tasks: applications, documents, transit coordination, data gathering. Left unchecked, they weigh down our care teams.
That’s why we’re using AI to automate the stuff that doesn’t require a human heart so our care guides have more time to use theirs. It pre-fills forms, finds available units, even helps with local transportation logistics. The time it saves is huge, but more importantly, it shifts the focus back to real, person-centered care.
Our best care guides aren’t great because they’re fast, they’re great because they notice when a voice cracks, when something’s off, and they step in. AI gives them more space to do that.
Of course, there’s a danger in swinging too far. When companies lean heavily into automation without considering the emotional and situational nuance of healthcare, people fall through the cracks.
AI doesn’t recognize when someone is crying while saying “I’m fine.” It can misinterpret tone, miss context, and overlook the unspoken cues that trained care guides are equipped to catch. In one recent case, a care team caught signs of potential suicidality during a phone call. Something no algorithm would have flagged.
That’s why any AI-driven program must prioritize seamless escalation to human support. There must be checkpoints, not just failsafes, where the system knows: this is where the human steps in.
What makes AI powerful isn’t just its speed, it’s its adaptability. It can play a lead role in one case and quietly stay in the background in another. That’s how it should be.
But no matter how we use it, success is measured the same way: in trust. And trust is fragile. It takes months to earn and seconds to lose. When we design AI to enhance, not replace human-first models, we keep that trust intact. Sometimes, we even strengthen it.
For health plan leaders and investors who feel overwhelmed by the flood of AI options: take a breath. Start with what already works. Look at programs that succeed with human-led models and ask where AI can quietly improve the margins, not uproot the foundation.
Work with partners who understand the space and who have a track record in human-first execution. Use AI to supplement, not supplant. The best results will come from integration, not disruption.
There’s enormous potential ahead. But the real win won’t be found in how futuristic a tool sounds. It will be measured in how many lives it quietly helps transform. One stable home, one trusted voice, one connected moment at a time.