Trust is not something we add to Healosphere after the fact. It is the first design question we ask, before any feature, before any interface decision, before any line of code.
That starts with understanding who we are building for. People who have been let down by health systems before are not being paranoid when they hesitate to hand over their health data to a new piece of technology. They are being reasonable.
What patients actually told us
In 2024, The Light Collective, in partnership with Dr. Maya Rockeymoore Cummings and a host of patients and community-based advocates, surveyed 377 patients, patient advocates, and caregivers about how they use digital health tools and what worries them about it. The results, published in the report “Tangled in the Web,” describe a real tension.
People are heavy users of digital health tools. Sixty-nine percent have used a wearable device to track a health condition, and seventy-three percent have taken part in an online health support group. But that comfort with technology sits right alongside real wariness. Seventy-four percent of respondents said they were moderately to extremely concerned about their personal health information being shared without consent. More than half said they would simply stop using an app altogether if it shared their data with advertisers.
The same pattern shows up around AI specifically. Most people (81%) are aware AI is already used in healthcare, and they see real upside: better disease screening, fewer diagnostic errors, more personalized treatment. But an overwhelming 91% said they want to be actively informed whenever AI plays a role in their care, and most believe the technology needs government oversight, not just a company’s word that it is being handled responsibly.
One finding stands out for how it reframes the whole conversation about data and consent. People are not simply “for” or “against” sharing their health data. Context decides everything. Only 29% said they would consent to share their data with a for-profit company, even one working on new treatments. But 64% would share the same data for research, and 58% would share it with a public health organization or nonprofit. People are not rejecting data sharing. They are rejecting being treated like a product.
What we heard directly from the people building healthcare
Alongside the published survey data, our own customer-discovery conversations, conducted through the NSF I-Corps program, surfaced the same tension from a different angle: the people delivering care every day. Per our practice of protecting the privacy of everyone we interview, the physicians, nurses, patients, hospital administrators, and researchers below are described by role only, not by name.
Physicians we spoke with described healthcare today as something closer to a self-service menu, where patients are expected to seek out care rather than have it brought to them. They pointed to a real data gap: home blood pressure monitoring exists, but the technology to get that data cleanly into a patient’s electronic medical record often does not, leaving nurses to collect readings manually over the phone. They also raised a harder problem underneath the data problem: most research behind today’s health algorithms comes from a narrow slice of the population, which means the tools built on it can perform worse for exactly the patients who need them most.
They were direct about what would actually change that. Not a “one-size-fits-all” tool, but one that understands that telling a patient to change their diet only works if the advice fits the food they actually eat, the language they speak, and the community they live in. Patients themselves echoed this: many expressed a strong preference for care from someone who shares their background, believing those providers better understand the toll chronic stress takes on the body over time. Cultural responsiveness, our interviewees told us, is not a nice-to-have feature. It is the feature that determines whether anyone uses the tool at all.
There was also a candid acknowledgment of the financial forces working against prevention. In a system that largely pays for sickness rather than health, keeping people well can actually cost a hospital revenue, even when everyone agrees it is the right thing to do. That misalignment is part of why AI-driven prevention tools face real headwinds getting adopted, not because the technology does not work, but because the incentives around it are pointed the wrong way.
What this means for how we build
Very different sources of evidence point to the same conclusion. People want the benefits digital health and AI can offer but want it to be secure, safe, and relevant for the context of their lives. They are also asking not to be an afterthought in how it is built, deployed, and explained to them.
That is the standard the HyperVigilant Healosphere is holding itself to: Would the people using this technology actually trust it and consider it relevant? And, would that trust and relevance be earned based on who was engaged and how it was built and deployed?
