AI-Linked Medicare Claims Are Surging. Here's How Healthcare Leaders Can Respond
Medicare claims tied to artificial intelligence have climbed sharply over the past seven years. A report from Manatt Health signals a clear takeaway: plan now for continued growth in AI-associated care.
If you manage clinical operations, revenue cycle, or compliance, this isn't a future problem. It's already shaping your volume, documentation standards, audits, and contracts.
Where the growth is showing up
- Imaging: AI-supported reads, triage, and decision support embedded in radiology workflows.
- Diagnostics: Cardiology, pathology, ophthalmology, and dermatology tools assisting with detection and risk scoring.
- Care management: Remote monitoring and virtual care with AI-driven alerts and prioritization.
- Documentation: Ambient scribe and coding aids that impact E/M levels and throughput.
- Operations: Prior auth screening, denials prediction, and capacity management.
What to do in the next 90 days
- Inventory AI use: List every tool touching clinical decisions, documentation, scheduling, or claims. Map each tool to service lines, CPT/HCPCS usage, and sites of care.
- Tighten documentation: Define what must be recorded when AI is used (who reviewed, final clinical judgment, tool name/version, rationale for acceptance or override).
- Update coding guidance: Clarify how AI outputs inform-but do not replace-medical decision-making. Train coders and clinicians together.
- Set model governance: Approve tools through a clinical, compliance, privacy, and security review. Require version control, change logs, and deactivation procedures.
- Vendor due diligence: Validate claims, review validation studies, FDA status if applicable, and indemnification. Require access to audit trails.
- Security check: Confirm HIPAA compliance, data flows, retention, and BAAs. Assess prompts and transcripts for PHI exposure.
- Patient messaging: Create plain-language explanations for how AI is used and who is responsible for care decisions.
Build guardrails for 2025
- Outcomes and ROI: Track time-to-diagnosis, read rates, false positives, length of stay, denial rates, and net revenue impact-per tool and per site.
- Equity and safety: Monitor performance by demographic groups. Escalate issues when accuracy drifts or workloads spike unexpectedly.
- Audit readiness: Preserve clinical oversight notes, model outputs, inputs, timestamps, and override reasons. Align with internal and external audit protocols.
- Coverage and prior auth: Keep a live matrix of payer rules for AI-associated services and any documentation requirements.
- Procurement standards: Require validation data, interoperability details, downtime plans, and exit clauses before signing.
- Sunset criteria: Define thresholds to pause or retire tools that fail safety, equity, or financial targets.
Compliance checkpoints
Expect closer scrutiny as volumes rise. Align your program with federal guidance and maintain proof that clinicians-not algorithms-make final decisions.
- Follow core program guidance from regulators, including escalation paths and board oversight.
- Keep full audit trails: inputs, outputs, timestamps, user IDs, and final sign-offs.
- Validate claims: Ensure marketing and clinical use match regulatory clearances where applicable.
- Manage model drift: Recheck performance after software updates or population shifts.
HHS-OIG compliance guidance and ONC's HTI-1 transparency requirements can help shape internal policy and oversight.
Data and integration essentials
AI only adds value if it fits cleanly into clinical and billing workflows. Build integration that reduces clicks, not adds them.
- Use structured outputs that flow into the EHR, PACS, and billing systems without manual copy/paste.
- Label AI-derived content clearly so clinicians can verify and coders can cite it correctly.
- Set owners for data quality, release management, and incident response.
People and skills
Your workforce needs fluency in AI-supported care: when to trust, when to question, and how to document the decision. Prioritize training for clinicians, coders, and revenue cycle teams.
If you need a fast way to upskill teams, explore job-focused options here: AI courses by job.
The bottom line
AI-linked Medicare claims are climbing and will keep climbing. Health systems that set governance, documentation, training, and measurement now will reduce audit risk, improve outcomes, and protect margins.
Don't wait for payers or auditors to define your standards. Define them yourself-and make them easy to follow at the point of care.
Your membership also unlocks:
