We Already Trust Technology With Our Sleep. We Just Don't Ask Anything of It.

We have quietly crossed a line we used to think we'd resist: we let technology into our sleep.

Tens of millions of people now wear a ring or a watch to bed, handing over the most private third of their lives to a sensor, and checking the verdict in the morning. Among insulin users specifically, about one in four already wears sleep tech to bed (DQ&A PWD Segmentation and JTBD Report, January 2026, Slide 57).

That is a real shift in trust.

A decade ago, the idea of a device watching you sleep felt invasive. Now it feels normal, even reassuring. The interesting question is not whether people trust technology with their sleep, it’s why we ask so little of it once we do.

Trust is measurable

In our new 2026 patient research, trust in automation isn't a vibe. It's a validated construct — a six-item composite with strong internal consistency (α=.91) — which means we can actually see how different people feel about handing a task to a machine.

What it shows is not universal trust, and I want to be honest about that. It shows that Luna's primary segments score at or above average on trust in automation and AI. Self-Sufficient Optimizers, about 1.2 million people, and High-Concern Strugglers, about 772,000, are the two groups most ready to let technology act on their behalf (DQ&A PWD Segmentation and JTBD Report, January 2026, Slides 5, 11, 14, 17, 62). Together those segments are roughly 2 million people — about 56.6% of the modeled US market on multiple daily injections or basal-only insulin with CGM, excluding pump users.¹²

The honest counterpoint sits in the same dataset. Skeptical Traditionalists, around 810,000 people, score lowest on trust in automation (Slide 15). This survey is a single snapshot, not evidence that trust has risen. Trust scores differ across the groups for which a sleep-only automation product is built. That's a more useful finding than a feel-good one, because it tells you who is ready and who isn't.

Willingness is not the same as adoption

There's a discipline worth holding onto here: attitude is not behavior. People telling a survey they trust automation is not the same as people using it every night. So I'll keep the willingness data in proportion: 76% of surveyed insulin users who track their sleep said they would definitely (44%) or likely (32%) adopt an app that integrates sleep, glucose, and insulin (slide 57).¹²

That is a strong intent signal. And it is still intent.

There's an older signal, too, and I'll flag it as such. Back in December 2021, a survey of 111 HCPs found that 84% were extremely or very interested in learning about a device providing overnight AID benefits for MDI patients, and 62% were extremely or very interested in the Luna concept before any clinical data existed (Seagrove HCP Quant, December 2021, n=111). That is nearly five years stale, and it predates the product. I include it only as direction: the appetite isn't new.

We trust the sensor. We haven't asked it to act.

Put it all together and a strange picture emerges: we have already decided, in large numbers, to trust technology with our sleep. We wear it. We believe what it reports. Then we adjust our day around its score and ask it to do nothing except tell us how we did.

That is the part that doesn't add up.

Sleep-only automation asks more of this existing trust: not just to watch the night, but to work in it. Automated insulin delivery while you sleep, stepping back when you wake.

The readiness is already here, in the segments built for it. The work is to earn it.

Nights protected. Days yours.

Sources

1. dQ&A, PWD Survey Results: Luna Patient Segmentation & Jobs To Be Done, January 2026 (fieldwork 13–19 December 2025), slides 5, 13 and 17 (modeled segment sizing), and slide 57 (app adoption intent among sleep monitors).

2. Seagrove Partners, Novel Device Concept – Quantitative Research (HCP), December 2021, n=111, pages 4–6 and 35. These are separate interest measures, not before/after results; this research is superseded by dQ&A HCP Quant, March 2026, n=317.