STAT Investigation Examines Commure’s Referral Payments And Customer Losses, As AI Health-Care Sales Scale

Commure’s pitch to health care providers is expansive: use artificial intelligence to automate administrative work, reduce the burden of bureaucracy, and improve the economics of care delivery. But a STAT investigation says the company’s sales strategy includes offering thousands of dollars in compensation to customers and other parties willing to refer its products to new business prospects, while some customers report steep financial losses and other negative outcomes after adopting its tools.

The company, currently valued at $7 billion, says its referral programs are industry standard and that the vast majority of its hundreds of customers are happy with its products and services. That tension between aggressive go-to-market tactics and uneven customer experience is the core signal in the report, because AI adoption in health care operations is increasingly being sold not as experimental software but as infrastructure that can directly affect the cost and quality of services received by millions of Americans.

The company’s reach

Based in Mountain View, Calif., Commure set out in 2020 to build what it described as a new operating system for health care. Its products target administrative functions including billing, scheduling, and clinical documentation, areas where provider organizations often hope automation can relieve staffing pressure and improve collections.

According to the STAT report, Commure claims its tools are used by more than 500 health care organizations across the country, including “130 of the nation’s largest health systems,” naming HCA Healthcare and Tenet Healthcare among them. Some customers have publicly described the company in highly favorable terms, saying in marketing videos and podcast interviews that its products restored joy and reliable revenue to their practices.

The scale of those claims is part of why the investigation matters commercially. Once software is embedded in revenue cycle and documentation workflows, changing course can be expensive and operationally disruptive, so sales practices and implementation outcomes become as important as the AI itself.

The sales model under scrutiny

STAT reported that internal company communications, legal filings, and customer contracts show Commure offers compensation in various forms to parties that help bring in new prospects. The publication said it reviewed documents and spoke with more than three dozen ex-employees, business partners, and current and former customers during a monthslong investigation.

CEO Tanay Tandon framed the company’s mission in a recent interview with Y Combinator as redistributing economic power in health care, saying he would like a world where the market cap of UnitedHealth is a fifth of its current level but every doctor is a millionaire. The investigation suggests the company is also using direct financial incentives to accelerate adoption of that vision.

On its own, a referral program is not proof of product weakness. But in health care, where software performance can affect revenue capture, staffing workflows, and patient-facing operations, incentive-heavy selling can make independent evaluation harder, particularly when the technology itself is changing quickly.

Why this matters for AI health-care adoption

The report’s broader implication is that the commercialization of AI in provider settings is entering a phase where distribution tactics deserve as much scrutiny as model performance. Health systems and clinics are not just buying automation; they are buying claims about financial return, workflow reliability, and operational risk.

Commure is among the most ambitious companies trying to automate the administrative layer of health care. The STAT investigation suggests that as these platforms scale, the market’s central question is shifting from whether AI can be inserted into billing and documentation to whether buyers can verify outcomes without being swayed by referral incentives, testimonials, or rapid sales expansion. In administrative AI, commercial viability depends not only on novelty or reach, but on whether promised savings persist once the software is live.