Commure, a technology company based in Mountain View, Calif., launched in 2020 with the stated aim of building a new operating system for health care. From its inception the company was organized around the idea of automation: using software and artificial intelligence to reduce administrative burden for clinicians. Company leaders cast the venture as a sort of populist technology effort to shift resources away from large insurers and toward practicing clinicians. In a public interview with an investor, Commure CEO Tanay Tandon described a vision in which doctors would see materially improved financial outcomes.
A central finding of the STAT investigation is that Commure offered substantial compensation — in many instances thousands of dollars — to customers and other parties who referred its products to prospective buyers. Internal documents and company communications reviewed by the reporters describe these payments as part of a deliberate strategy to accelerate sales. Commure has characterized such referral programs as common in the industry; nevertheless, the investigation highlights the scale and centrality of these financial incentives to the company’s go-to-market approach.
While many customers provide glowing testimonials in marketing materials and interviews — saying the products restored joy and delivered reliable revenue — the STAT reporting identified a contrasting set of accounts. Some clinics and customers told investigators that after adopting Commure’s tools they experienced steep financial losses and other negative outcomes. Those reports were collected through interviews with current and former customers as well as review of customer contracts and internal communications.
The company disputes that these adverse outcomes are representative. Commure maintains that the vast majority of its hundreds of customers are satisfied with its products and services. Specific details about the clinics that reported financial harm, the amounts involved, or remedial measures were not fully disclosed in the portions of the investigation available without subscription.
Commure publicly states that its products are used by more than 500 health care organizations and by “130 of the nation’s largest health systems.” The company is reported to have enterprise customers that include large, for-profit health systems such as HCA Healthcare and Tenet Healthcare. In marketing and promotional materials, customers who support Commure’s technology describe improvements in billing, scheduling, and clinical documentation.
At the same time, the company’s compensation-based referral strategy is presented internally and externally as a tool to rapidly expand adoption of AI-driven administrative software across diverse clinical settings.
The STAT investigation drew on multiple sources: internal company communications, legal filings, customer contracts, and interviews. The reporters said they spoke with more than three dozen former employees, business partners, and current and former customers while compiling the story. STAT published a highlights summary of the findings; readers are informed that the remainder of the in-depth reporting is available behind a STAT+ subscription paywall.
The investigation situates Commure’s trajectory within a broader debate about the rapid commercialization of AI in health care. The company’s attempt to sell AI-based automation quickly and widely — combined with substantial financial incentives for referrals and customization for different clinical settings — intensifies challenges in evaluating products that affect both the cost and quality of health services. The reporting underscores that AI is a constantly evolving and often opaque technology, and that rapid deployment across heterogeneous clinical environments can produce unequal results for different customers.
Taken together, the findings in the STAT highlights illustrate tensions between ambitious promises of automation and real-world implementation risks. The pieces of the full investigation that detail company documents, specific customer allegations, and internal deliberations are reported by STAT as part of a longer subscriber-only article; these underlying materials were the sources for the summary findings reported above.