Burda on Healthcare: Following the Healthcare AI Money Leads to One Inescapable Conclusion

Follow the money, right? Whether that’s proceeds from the sale of stolen Venezuelan oil flowing into offshore accounts controlled by a convicted felon or capital flowing into artificial intelligence-powered healthcare technologies. Money reveals motive.

With that in mind, let’s take a look at a handful of recent reports, studies, surveys and research on where healthcare organizations are spending their healthcare AI dollars. Collectively, they might reveal motive.

Streamlining Administrative Operations

Last August, Sage Growth Partners, a Baltimore-based healthcare consulting firm, released an 11-page report called The Healthcare C-Suite’s Take on AI. The report is based on a survey of 101 hospital and health system “C-suite leaders.” Here are some of the more noteworthy findings:

• 50% of the respondents said their organizations are “actively exploring AI opportunities” or “aggressively pursuing AI solutions.”
• 67% said their organizations are investing in AI technologies to “enhance patient care,” while nearly the same share — 66% — said they are doing so to “streamline administrative operations.”
• 77% said AI can help most in improving revenue cycle operations, second only to expanding home and remote patient monitoring initiatives, cited by 81% of the respondents.

“Executives see incredible potential in AI for advancing top strategic initiatives. From revenue cycle operations, integrating virtual care into traditional delivery models, and overall digital transformation, C-suite leaders are expecting AI to have a substantial impact in driving core objectives,” the Sage Growth Partners report said.

Facilitating Scheduling and Simplifying Billing

Last September, the National Coordinator for Health Information Technology/Assistant Secretary for Technology Policy at HHS released a 17-page report called Hospital Trends in the Use, Evaluation, and Governance of Predictive AI. The report is based on data from the American Hospital Association (AHA) surveys of hospitals in 2023 and 2024. The AHA surveyed 2,547 hospitals in 2023 and 2,253 hospitals in 2024.

Here are some of the more noteworthy findings:

• 71% of the hospitals said they used predictive AI in 2024, up from 66% of hospitals that said the same thing in 2023.
• 92% of the hospitals, both years, said they used predictive AI to predict health trajectories or risks for inpatients.
• 67% of the 2024 hospitals said they used predictive AI to facilitate scheduling compared with 51% of the 2023 hospitals — a jump of 16 percentage points.
• 61% of the 2024 group said they used predictive AI to simplify billing compared with 36% of the 2023 hospitals — a jump of 25 percentage points.

“The fastest growing use cases for predictive AI were billing simplification and scheduling facilitation,” the HHS report said.

Improving Efficiencies and Margins

Last October, Menlo Ventures, a private venture capital firm based in Menlo Park, California, released an 18-page report called 2025: The State of AI in Healthcare. The report is based on a survey of more than 700 senior executives involved in AI decision-making in three healthcare industry sectors: providers, insurers and pharma/biotech. Here are some of the more noteworthy findings:

• 22% of the healthcare organizations across all three sectors have licensed commercial AI applications. Health systems topped all sectors at 27%. Payers were at 14%.
• Of the $1.4 billion represented by those commercial AI application licenses, health systems spent a little more than $1 billion or 75% of the total.
• Health systems spent most of their money on two use cases: ambient scribes ($475 million) and coding/billing ($400 million).
• Patient engagement ($70 million), prior authorization ($50 million) and payer engagement ($50 million) were a distant third, fourth and fifth, respectively.

“AI agents offer the promise of improving efficiency and margins without compromising care quality,” the report said.

Prioritizing Revenue Cycle Management

Also in October, Bain & Company, a Boston-based management consulting firm, released a report called Healthcare IT Investment: AI Moves From Pilot to Production. The report is based on a survey of 228 U.S. healthcare provider and payer executives. Here are some of the more noteworthy findings:

• 49% of the provider respondents cited revenue cycle management (RCM) as one of their top three investment priorities in 2025, topping a list of 10 choices.
• 62% of the provider respondents cited documentation support (scribe/ambient listening) as their top use case for AI, followed by clinical documentation improvement and compliance assurance for payer interactions (43%), medical coding (30%) and prior authorizations (27%).

“RCM’s appeal lies in hard‑dollar ROI: Accurate documentation and coding, resulting in cleaner claims and fewer denials, lead to measurable gains on both the revenue and expense lines,” the Bain report said.

Leveraging Customer Relations to Proactively Schedule Care

Last November, Chartis, a Chicago-based healthcare consulting firm, released its latest annual healthcare digital transformation report. The report is based on a survey of 150 health system executives. Here are some of the more noteworthy findings:

• 91% of the respondents said it was “very” or “somewhat” important to leverage customer relations management systems to proactively schedule care.
• 91% said it was “very” or “somewhat” important to add/expand AI-supported capacity and referral management.
• 87% and 86%, respectively, said it was “very” or “somewhat” important to use AI health coaches to answer questions and help patients follow their care plans and use AI to predict health risks.

“Executives expect digital and AI tools will allow their health systems to better anticipate and meet demand,” the Chartis report said.

Enabling Alternative Payment Model Success

In December, JAMA Network Open published a study called Uptake of Generative AI Integrated With Electronic Health Records in U.S. Hospitals. Three health services researchers affiliated with the Office of the National Coordinator for Health Information Technology/Assistant Secretary for Technology Policy at HHS and the University of Minnesota School of Public Health did the study. The study is based on an American Hospital Association survey of 2,174 hospitals.

The survey asked hospitals about the integration of generative AI technology into their EHR systems. The researchers divided hospitals into three buckets: early adopters (they have already integrated generative AI into EHRs); fast followers (they intend to do so with the next 12 months); and delayed adopters (they intend to do so within the next five years, have no current plans to do so or weren’t sure).

Here’s how the hospitals broke out:

• 43.7% were delayed adopters
• 31.5% were early adopters
• 24.7% were fast followers

The researchers then broke the results down by different variables like size, ownership, margin, location, etc. The most revealing variable was the number of alternative payment models. Hospitals that participated in more alternative payment models, which pay them based on outcomes rather than volume, tended to be early adopters of generative AI in their EHR systems. To wit, 55% of hospitals that participated in at least four alternative payment models were early adopters compared with 26.3% of fast followers and 18.7% of delayed adopters.

Improving Revenue Capture Performance

Also in December, KLAS Research released a report entitled Healthcare AI Update 2025. The seven-page report is based on a survey of 3,370 executives from 1,742 healthcare organizations of different stripes. KLAS refers to the report as a “perception study” that captures how the executives are using AI now and how they plan on using it in the near future in seven use case categories.

Today, the top use case category was “clinical,” but not clinical as in diagnosis and treatment. It’s clinical as in ambient speech, cited by 79% of the respondents. Revenue cycle management/claims was fourth on the list of seven categories with claims adjudication, coding automation and denials management as the specific use cases in that category.

But what about the near future? Revenue cycle management/claims jumped into the top slot of use case categories, cited by 53% of the respondents. Clinical dropped down to third at 43% with patient/member engagement sandwiched in between at 48%.

“Most AI use cases center on clinical and administrative workflows, but organizations are increasingly expanding into revenue cycle, where many organizations hope to improve revenue capture amid shifting reimbursement requirements being pushed by the U.S. government,” the report said.

Increasing Physician Productivity

Speaking of ambient AI-powered medical scribes, a new study published in January in JAMA Network Open suggested the reasons why they’re currently the most popular AI technology being deployed by healthcare organizations, as noted in the KLAS perception study.

The study, conducted by 10 researchers affiliated with the University of California San Francisco (UCSF), examined the connection between the use of ambient AI scribes by physicians and several measures of physician productivity. The study pool consisted of about 1.2 million ambulatory care visits by patients to the UCSF health system from Jan. 1, 2023, through April 1, 2025, and the 1,565 physicians who treated them. About 183,000 visits were with about 700 physicians who used ambient AI scribes.

Compared with the physicians who didn’t use ambient AI scribes, the physicians who did provided more intensive care per visit and per week, as measured by relative value units (RVUs) and saw more patients per week. All with no increase in claim denials from health plans. In terms of dollars, based on the most current Medicare physician fee schedule, the use of ambient AI scribes generated more than $3,000 in extra payment per physician per year.

“Further research should assess factors in AI scribe use, ascertain whether increased RVUs reflect more clinical services or accurate coding rather than upcoding,” the researchers said.

In other words, the use of ambient AI scribes is more about charge capture and patient volume than it is about reducing the clinical documentation burden on doctors so they have more face time with patients.

One Inescapable Conclusion

You don’t need agentic AI technology to figure out what’s going on. Healthcare executives have figured out that AI can be of most use to their respective organizations in revenue cycle management, or as we once called it, billing and collections. That’s where the return on investment is, and that’s where they’re increasingly putting their AI dollars.

David W. Johnson, the founder and CEO of 4sight Health, likes to say that the smartest people in healthcare work in revenue cycle. Give them the smartest technology, and the average person with a nominal copay doesn’t stand a chance.

Just think if we shifted all that spending on AI into wellness and prevention, we wouldn’t have to spend it on robots chasing down sick and injured patients for their last dime.

Let’s build a better healthcare system.