AI Medical Scribes Drive Up Hospital Bills Without Changing Care

AI documentation inflates medical coding

Artificial intelligence tools designed to assist with clinical documentation are contributing to rising healthcare costs by capturing more billable details than human clinicians typically record, according to a new analysis from PwC. The consultancy identifies AI as one of five factors pushing health costs up to 9% in 2027—matching 2026's rate, the highest since 2010–11.

The mechanism is straightforward: AI-powered note-taking systems document granular specifics about patient conditions and complications that time-pressed physicians might consolidate under a single diagnostic code. These additional details can justify assigning higher-severity billing codes to insurers, even when the actual medical care delivered remains unchanged.

The maternity ward case study

Blue Cross Blue Shield examined this phenomenon in hospital maternity departments and found stark evidence of the pattern. Between 2022 and 2025, some facilities saw diagnoses of acute posthemorrhagic anemia in new mothers jump from 4% to 12.3% of admissions. Yet blood transfusions—the standard treatment for this condition—remained essentially flat during the same period.

An audit of the hospital system showing the steepest increase revealed that fewer than one in five cases coded for this diagnosis actually met clinical criteria. Blue Cross Blue Shield directly linked this "coding intensity" surge to hospitals' adoption of AI billing systems, calculating that it added $22 million to maternity spending across the studied facilities in just three years.

Why it matters

This represents a fundamental tension in healthcare AI deployment. While the technology promises long-term efficiency gains through administrative automation and earlier disease detection, its first measurable impact has been to optimize revenue extraction rather than cost reduction. The pattern illustrates how AI systems amplify existing incentives within their deployment environment—in this case, a fee-for-service model that rewards more detailed documentation of patient complexity.

Context and caveats

One of the PwC report authors told Healthcare Dive that AI remains a secondary cost driver compared to traditional factors like labor and supply expenses. The technology's role could shift over time as hospitals deploy AI for broader administrative functions or clinical decision support.

The dynamic reflects what one health insurance executive described as the predictable application of any new tool: "Companies will take AI and say, 'How can I use this to further my self-interest?'"

These findings were originally reported by Tech Brew, citing the PwC healthcare cost analysis and Blue Cross Blue Shield data.