AI is already making health care more expensive, and it's probably only going to get worse.

Why it matters: For all of the ways AI could meaningfully improve patients' lives, making care even more unaffordable isn't one of them.

Driving the news: PwC yesterday estimated that medical costs will go up by 9% in the employer market next year, and by 8.5% in the individual market. One of the largest drivers is providers' use of AI-enabled software and scribes that more thoroughly document the care that's delivered.

• Such tools are being used "to capture greater billing complexity, and plans are absorbing the cost," per the report.
• PwC said the financial impact isn't so much due to people using more medical services as "changes in coded severity, case mix and paid amount per claim."

The big picture: Within the health care system, the current incentives are "to do more and get paid for more," Paul Markovich, CEO of Blue Shield of California parent company Ascendiun, told me this week.

• Companies "will take AI and say, 'How can I use this to further my self-interest?'" he said.
• However, he added that AI "ultimately will bring a lot of the administrative costs out of the system."

Between the lines: It's not only that AI is helping providers make more money per unit of service. It's also poised to flood the system with more products and services.

• "AI makes any system more efficient — and since our health system is already super efficient at driving fee-for-service units of care and coding, I think it is going to drive up both and make health care spending grow even faster," Venrock partner Bob Kocher told me.
• Most of the investment and adoption is for managing revenue cycles and for drug development that will bring promising but pricey new drugs to market, he added.

Yes, but: The current hype around using AI for administrative tasks — including billing — will burn off, shifting the focus to AI uses that improve patient outcomes, Harvard Medical School associate professor Hossein Estiri said.

• "I think health care systems are beginning to realize that the market narrative isn't where the real value is. I think the real value is to improve patient lives," he told me.
• AI will "enable more proactive health and make care more precise," he added. The result is fewer sick people and a lower cost of care.

What we're watching: Whether AI's arrival changes incentives along with payment systems.

• "Efforts to drive more non-fee-for-service payment models that reward lower total costs will drive adoption of AI to achieve these goals," Kocher said.
• Value-based care — essentially paying for outcomes instead of per unit of care — is an idea that's been thrown around forever. AI could be the thing that truly makes it a necessity.

A timely new UBS report analyzes the impact of AI tools on both insurers and hospitals through financial and competitive lenses.

The big picture: Both sectors are adopting AI for billing and claims, reducing costs and creating efficiencies.

• But in the long term, AI will probably increase the dominance of the country's largest health systems, the report argues, which is different from how it will impact the insurer market.

AI will likely make the entire insurance industry more efficient, but the financial gains from lower administrative costs will be "competed away over time" and reinvested in other ways.

• That's because administrative AI use isn't likely to give any one insurer an advantage, with the entire industry pursuing the same efficiencies.

Among hospitals, for-profit operators like HCA, Tenet and UHS have the financial and operational ability to invest aggressively in AI faster than nonprofit hospitals.

• That market advantage may last for a while.
• "AI has the potential to create a more persistent competitive gap in hospitals, as early movers not only realize immediate efficiency gains but also build institutional capabilities and data advantages that may take slower-moving peers years to replicate," UBS argues.

The bottom line: Bigger is better, the note concludes.

• "While AI adoption will be widespread across the healthcare ecosystem, the most meaningful near-term economic benefits are likely to remain concentrated among the largest and most well capitalized organizations."