Today the New York Times published a piece that confirms what anyone in medical billing already knows: insurers are using AI to deny claims at industrial scale, and hospitals are using AI to bill bigger. The only people without AI on their side are the ones who need it most — independent practices.
The numbers are staggering
According to the Blue Cross Blue Shield Association's own analysis, hospitals' use of AI coding tools generated nearly $1 billion in additional expenses over just two years. One Michigan health system, McLaren Health Care, reported adding $1 million per month in revenue after deploying AI from SmarterDx — a tool that suggests additional diagnoses and documentation to justify higher payments.
On the other side, insurers have deployed AI to scan charts and auto-deny claims. Prior authorization systems increasingly rely on algorithms to determine whether care should be covered. Benefit consultants cited in the article predict employer health costs could rise as much as 11 percent next year, with AI-driven billing disputes as a contributing factor.
"Bots fighting bots" — but only for the big players
The article quotes Dr. Shiv Rao, a cardiologist and founder of Abridge, warning about "a horrible dystopic future" of "bots fighting bots, agents fighting agents." Dr. Joon S. Lee, CEO of Emory Healthcare, described his nightmare scenario: "A whole data center could be occupied just doing that."
But here's what the article doesn't say: this AI arms race is between hospital systems and insurance companies. Both sides have billion-dollar budgets. Both are deploying sophisticated technology. Independent practices — the ones actually delivering care in communities across the country — have neither.
McLaren has SmarterDx. Blue Cross has Basys.ai. Cigna has its own AI prior auth systems. A five-physician practice in Tampa has a billing manager with a spreadsheet and a phone.
Upcoding vs. recovering what you're owed
The article raises legitimate concerns about AI being used to inflate claims. Hospitals are adding secondary diagnoses — anemia, low sodium — that justify higher payments without evidence of different treatment. That's a real problem, and the Blue Cross Association is right to flag it.
But there's a critical distinction the article only brushes against: there's a difference between using AI to bill more than you're owed and using AI to recover what you're already owed.
Independent practices aren't looking to upcode. They're looking to stop losing money on claims that were correctly billed, legitimately provided, and wrongly denied. The average practice has a denial rate between 5% and 15%. At an average claim value of $250, a 500-claim-per-month practice is leaving $6,000 to $19,000 on the table every single month — not because the care wasn't warranted, but because the appeal process is too slow, too manual, and too expensive to fight every denial.
That's the gap apruvl fills. We don't suggest new diagnoses. We don't inflate claims. We analyze your remittances, identify denials that should be appealed, and generate evidence-based appeal letters grounded in your actual clinical documentation and the payer's own policies. Every appeal is reviewed by your team before it goes out the door.
The real cost of doing nothing
Harvard health economist David Cutler, quoted in the article, suggested AI might "force a difficult conversation about how doctors and hospitals are paid." He's right. But that conversation won't include independent practices unless they have a seat at the table — and right now, the table is set for organizations with seven-figure technology budgets.
The article also quotes Caroline Pearson of the Peterson Health Technology Institute, noting that both sides are willing to go extra rounds on claims disputes "because it's cheap." Cheap for a hospital system or a national insurer, maybe. For a four-person billing team at an orthopedic practice, every hour spent on appeals is an hour not spent on collections, eligibility checks, or patient calls.
AI doesn't have to mean bots fighting bots. It can mean giving your billing manager the same analytical power that hospital systems and insurance companies already have — without the enterprise price tag, without the implementation project, and without the compliance risk.
What this means for your practice
If you're running an independent practice, today's article should be a wake-up call, not a reason to panic. Here's the takeaway:
- Insurers are using AI to deny faster. Your appeals need to be faster too. Manual processes can't keep up with automated denials.
- Hospitals are using AI to document more aggressively. You don't need to follow their playbook — you need to make sure your legitimate claims aren't falling through the cracks.
- The cost of fighting denials is dropping. AI tools designed for independent practices are now available at a fraction of what hospital systems pay. The economics of appealing a $300 denial just changed.
- Doing nothing is more expensive than ever. As insurer AI gets more sophisticated, denial rates will climb. Practices that don't adapt will lose more revenue every year.
See what you're leaving on the table. Upload your claims file and get a free audit in 60 seconds. No credit card. No commitment. Just a clear picture of your recoverable revenue.