AI is helping healthcare providers and insurers process more work faster, but it may also be quietly driving up costs across the system. A new report from the Peterson Health Technology Institute finds that while AI is reducing administrative burden at the organization level, it is not lowering overall system costs and may actually be increasing them as more transactions, billing activity, and prior authorization requests flood the system.
Here’s what matters for organizations working with healthcare providers and medical debt:
The big picture
- The U.S. healthcare system carries roughly $350 billion in annual administrative waste.
- About $266 billion comes from administrative complexity, with another $59 billion to $84 billion tied to fraud and abuse.
- AI adoption is accelerating across two key areas: prior authorization and medical billing.
What’s happening in prior authorization
- Providers are using AI to automate submissions and appeals.
- Health plans are using AI to triage and evaluate requests.
- The result: more volume, not necessarily more efficiency.
One participant described it bluntly: AI tools are creating “bot wars,” where automated systems on both sides generate more back-and-forth instead of resolving requests faster.
Key takeaway:
- AI may lower the cost per transaction for individual organizations, but total system activity is increasing, which offsets those gains.
What’s happening in medical billing
- AI-powered documentation tools, including ambient scribes, are enabling more detailed coding.
- That leads to higher billing intensity, even when care delivery hasn’t changed.
The report highlights:
- A measurable increase in higher-complexity billing codes after AI adoption.
- Revenue gains for providers tied to more complete documentation.
Health plans are already responding:
- Deploying AI to identify and adjust higher-cost claims
- Implementing downcoding and reimbursement reductions
The long-term impact of these responses is still unclear.
Why this matters for collections and healthcare providers
- Higher billing amounts can translate into larger patient balances
- Increased back-and-forth between payers and providers can delay payment cycles
- More complexity upstream often leads to more friction downstream in collections
The bottom line: AI is making it easier to do administrative work, but it is not fixing the underlying systems that create complexity in the first place. Without changes to workflows, reimbursement models, and policy structures, the technology may continue to increase activity and costs rather than reduce them.
