AI Health Momentum: Function Health's $298M Raise And Why Agentic Payments Matter For Healthcare
Function Health has raised $298 million to scale AI-centered health testing and longitudinal care. The company's pitch is straightforward: routine whole-body labs, advanced imaging, and access to physicians-wrapped with AI-so members can "pursue 100 healthy years," as one investor described it.
Why this funding matters for healthcare leaders
The new round reportedly puts Function's valuation at $2.8 billion and arrives alongside the debut of Medical Intelligence Lab, a generative AI model trained by physicians to provide personalized insights from a member's data, content and current research. Function is also rolling out an AI assistant that can reference prior labs, physician notes, and scans to answer questions and guide next steps.
As CEO Jonathan Swerdlin put it: "It is not good enough to be in a world where AI exists and not be applying it to your health. You should be able to manage your biology. The objective of Function Health is to apply the best available technology to human health."
The broader signal: funding is back, evidence still rules
Healthcare AI is attracting serious capital again. In the first half of the year, healthtech venture funding rebounded to $7.9 billion, with AI-heavy rounds like Ambience Healthcare's $243 million. Big tech is leaning in-Amazon and Nvidia are eyeing medical imaging and diagnostics-while governments move to shape guardrails. The U.K. is working on oversight for AI in care, and a U.S. executive order earmarked $50 million for AI-driven pediatric cancer research.
Policy pilots are underway too: Medicare is preparing an AI-assisted prior authorization test in six states to see if decisions can speed up without sacrificing safety. Still, caution is warranted. Some high-profile models-for example, attempts to predict genetic mutations from pathology slides-have shown sensitivities near 60 percent, which slows clinical trust and adoption. The takeaway: budgets are flowing, but deployment hinges on results that stand up to real-world evidence.
What providers and payers can do now
- Start narrow: pick 1-2 use cases with clear metrics (TAT reduction, readmission risk, cost per episode).
- Get your data house in order: stable identifiers, lab normalization, imaging access, consent artifacts, audit trails.
- Stand up clinical governance: bias checks, failure modes, escalation policies, model versioning, and sunset plans.
- Integrate where care happens: EHR hooks, PACS access, secure patient messaging, clinician-in-the-loop workflows.
- Demand proof: external validation, prospective pilots, and outcome deltas-not just ROC curves.
- Budget for change management: training, documentation, and feedback channels for clinicians and patients.
Payments are going "agentic" - this touches revenue cycle, patient pay, and digital care
EMVCo, the body behind EMV Specifications, is exploring how EMV 3-D Secure, Payment Tokenisation, and Secure Remote Commerce could support agent-led card payments. Agentic payments change how transactions are initiated, authenticated, and secured, so global, interoperable specs matter if you want trust at scale. EMVCo (backed by American Express, Discover, JCB, Mastercard, UnionPay, and Visa) is inviting stakeholders to weigh in.
Why this matters in healthcare: as AI agents begin scheduling visits, ordering tests, and managing recurring payments for memberships or remote monitoring, you'll see more agent-initiated transactions. That impacts identity, consent, chargebacks, and fraud controls. Getting ahead of standards now can reduce friction and risk later.
- Align with modern rails: tokenization for card-on-file, 3DS for step-up when risk rises, and SRC for consistent checkout.
- Design consent once, use everywhere: agent permissions, payment limits, and revocation flows that patients can understand.
- Instrument the stack: event-level logs that link clinical agent actions to payment events for audit and disputes.
- Run "edge case" drills: expired cards, partial authorizations, prior auth changes, caregiver vs. patient identity.
- Plug into standards work: track EMVCo updates and monitor open approaches like Google's Agent Payments Protocol (AP2) and the Agentic Commerce Protocol.
EMVCo is the best place to follow specification progress. For a view on agent-led payments from the platform side, see Google's Agent Payments Protocol (AP2).
If you're upskilling teams on practical AI
If your clinical ops, RCM, or digital teams need a structured way to get AI-literate fast, you can browse role-based options here: Complete AI Training - Courses by Job.
Bottom line
Function's raise signals sustained momentum for AI in care delivery, but adoption will follow evidence and trust. Build the proof, lock down data and consent, and prepare your payments stack for agent-led flows. Do that, and you'll be ready to turn AI from a headline into measurable patient and business outcomes.
Your membership also unlocks:
