The J.P. Morgan Healthcare Conference sets the mood for the rest of the year. It tells you what digital health money is going to chase, what it's nervous about, and what it has quietly given up on. This year the room had a specific feeling: controlled buzz. People were excited, but they'd been burned enough times to keep one hand on their wallet. Six themes came up in nearly every meeting I sat in.
DTC came back from the dead
Direct-to-consumer healthcare has been a money pit for a decade. Google, Amazon, and Apple all took serious swings at it, and all of them pulled back. So it was strange to watch the energy return this fast. The trigger was the wave of conversational health products built on large language models that Anthropic and OpenAI shipped over the past few weeks. A category everyone had written off suddenly looked alive again.
The excitement was about the implication more than the chatbots themselves. If the foundation models can hold a useful health conversation, a whole tier of third-party companies gets built on top of them, the same way a tier of companies got built on top of the app store. That second-order opportunity is what had investors leaning forward.
M&A, not venture, is the story this year
Ask a dealmaker what moves the most capital in 2026 and they won't say venture rounds. They'll say acquisitions. The market is consolidating, and the buyers know it.
Picking the right target is the easy half. Operators underestimate the rest. The hard part is absorption: can the acquirer actually fold the company in without the integration eating the value it paid for? I heard more conversations about post-close integration than about valuation. That tells you where the pain has been.
The AI conversation grew up
AI is still the hottest thing in the building, but the questions changed. A year ago people asked what AI could theoretically do. This year they asked whether it works inside a real clinical or operational setting, and most of the answers came back to data.
An AI-first platform lives or dies on four things: clean data, redesigned workflows, real-time interoperability, and infrastructure that holds up as volume grows. Miss any of them and the value stalls, no matter how good the model is. The data advantage is real, and it's specific. Clinical data, operational data, whatever you've got works far better once it's activated inside the buyer's own environment. Data inside the walls beats data you're trying to reach across them. Which is part of why M&A matters so much this year.
Payers are bracing for a bad year
The unwinding of the safety net and the pressure on Medicaid is the regulatory headache nobody at the conference wanted and everybody had to plan around. The reimbursement losses are already forcing cost to shift somewhere, and it's landing on payers at the same time three other lines keep climbing: high-cost medical utilization, pharmacy spend driven by GLP-1 demand, and mental health.
So payers are doing the unglamorous work. Tighter utilization management. Hard limits on technology and operational spend. The posture in those rooms was defensive, and it should be.
Employer health benefits are running out of road
Medical cost inflation is hitting employer plans somewhere between 9% and 25%, and it's squeezing the whole supplemental benefits budget. This pressure runs straight through 2026.
Tweaking the existing plan won't bend that curve, and everyone in those meetings knew it. Employers are starting to look at structurally different ways to cover people: ICHRA, alternative payers, going direct to employer, third-party administrators bringing their own networks. The companies that help employers make that switch cleanly have a real opening. The ones selling another round of plan-design tweaks are selling aspirin for a broken leg.
Longevity is population health with better funding
Every other booth seemed to have a longevity angle. Predict your cardiovascular age, extend your healthspan, the whole pitch. Strip away the language and it's prevention, which is the thing population health was supposed to deliver and mostly didn't.
Which is fine by me. If a glossier, better-capitalized version of prevention is what finally gets this country to spend money before people get sick instead of after, I'll take it. Longevity might be the trojan horse that smuggles real preventive investment in through the front door. Worth watching whether the money actually follows the story.
The mood overall: people want to believe again, and for the first time in a while they have a reason to. Whether it holds depends on whether the AI products do real work and whether the buyers can absorb what they're acquiring. Both are open questions. Ask me again in twelve months.d faster to no. Build for both.