The denial arrives just three hours after the submission is made. Its language echoes the diagnosis codes rather than the clinical chart itself. The letter makes no mention of the specific medical necessity criterion used to justify the decision, and nowhere on the document is there a clinician who has signed off or put their license on the line. This experience is becoming alarmingly common.
A staggering three-quarters of physicians report that denials have increased over the past five years. According to the American Medical Association’s 2026 prior authorization survey, six in ten physicians now worry that artificial intelligence (AI) is at the root of the growing number of denials. The prior authorization process alone consumes an average of 13 hours each week from physicians and their staff, time that could otherwise be spent caring for patients. In fact, two out of every five medical practices now employ at least one staff member whose sole responsibility is managing prior authorizations and denials.
Despite these challenges, the leverage available to practices has improved compared to the previous year. In 2026, seven states enacted new laws requiring that any adverse determination regarding medical necessity must be made by a licensed clinician. Additionally, under the CMS interoperability rule, payers are now required to publicly post their own prior authorization metrics, including the percentage of denials that are overturned on appeal. Practices that excel in this new landscape are those that treat an automated denial not as a dead end, but as a documentation challenge to be tackled strategically. Below are eight actionable strategies to optimize your approach.
1. Recognize the Signs of an Algorithmic Denial
Automated denials have distinctive characteristics. They are returned with remarkable speed—often within mere hours of submission. The rationale provided is typically generic, mirroring the codes on the claim rather than responding to the specific details within the physician’s note. No reviewing clinician is identified, and the denial letter refers to a policy in vague, general terms without pinpointing the precise criterion that was not met.
It is crucial to train your billing staff to recognize and flag these automated denials separately from those that are manually reviewed. The appeal process for an automated, algorithm-driven denial is fundamentally different than for one that a human clinician has reviewed. If a denial results from a human error in clinical judgment, the appeal is a clinical argument. When no human has reviewed the claim, the appeal should focus on the process—and process-based appeals often win faster. Practices should track these two types of denials separately in their management system, ensuring that volumes are visible when renegotiating contracts or filing complaints with the state insurance department.
2. Request the Reviewer’s Name, Credentials, and Specialty in Writing
Every time a denial is received, send a written request (never just a phone call) asking for three specific items:
- The name of the individual who reviewed the claim
- The license they hold
- Whether they practice in the specialty that manages the patient’s condition
Most state utilization review statutes require insurance plans to provide this information upon request. The answer you receive will clarify your next steps. If a qualified physician reviewer is named, you now have someone with whom you can schedule a peer-to-peer discussion. If the insurer cannot provide a name, you have documented evidence that no licensed clinician reviewed the denial—a key fact that can sway an appeal in your favor and, in many states, is now the deciding issue for overturning a denial.
To streamline this process, create a template for these requests so that they are automatically sent for every denial flagged as automated. This ensures consistency and prevents the process from being overlooked.
3. Reference Your State’s AI Law by Section Number in Appeals
At least seven states passed laws in 2026 that restrict the use of AI by payers, with several more already having similar regulations on the books. For example, Washington’s Senate Bill 5395, which took effect in June 2026, stipulates that only a licensed physician or health professional may deny a prior authorization request on the grounds of medical necessity. The law also bars insurers from relying solely on AI for these decisions. Alabama, Colorado, and Georgia have passed similar measures, while California and Maryland had such requirements already in place since 2025.
When submitting an appeal, research the specific statute for your state and payer type. Be sure to cite the law and include the relevant section number in your appeal letter. Appeals that reference a concrete legal obligation are routed differently than those that simply argue on clinical merit; compliance concerns carry more weight with those reviewing the case. However, be mindful of the limitation: self-funded employer plans generally fall outside the scope of state insurance regulations, so check your contracts to ensure your argument applies.
4. Leverage the Payer’s Own Denial Data
Since 2026, Medicare Advantage plans, Medicaid managed care plans, and qualified health plan issuers have been required to publish annual prior authorization metrics on a public website. These statistics include:
- The percentage of requests approved
- The percentage denied
- The share approved after appeal
- The average time between submission and decision
These public metrics are invaluable when contesting a denial. If a plan reports that a significant percentage of its denials are overturned on appeal, this is evidence that its first-pass review process is unreliable. This fact should be included in your appeal letter and raised during contract negotiations. Additionally, pay close attention to the average decision time. If a plan is rendering determinations in a fraction of the time outlined in its policy, that strongly suggests the use of an automated system, regardless of whether they admit to using AI.
5. Build and Analyze Your Own Denial Data in Advance
Don’t wait until you face a crisis to gather data. Run monthly reports by payer, CPT code, and turnaround time from submission to determination. Many practices track only their overall denial rate, which signals that something is wrong but lacks the detail needed to prove it.
Patterns in your data are powerful evidence. For instance, if you receive 20 denials for the same CPT code from the same payer, all within four hours and using the same generic language, you have documented a systemic problem worthy of escalation to the state insurance commissioner or the insurer’s provider relations team. One isolated denial is an anecdote; a pattern is an argument. Most practice management systems can generate these reports with a saved query. Ensure the process for creating and maintaining these reports is clearly documented to avoid losing this capability if a key staff member leaves.
6. Structure Your Notes So Algorithms Can Find the Necessary Criteria
When an algorithm reviews a claim, it matches structured data in the submission against the payer’s policy criteria. It doesn’t interpret narrative prose for nuance or context. If your justification for a service appears only in the free-text assessment, the algorithm will miss it.
To address this, extract the medical necessity criteria from the payer’s policy and ensure these elements are explicitly documented in structured fields. Key elements include:
- Diagnosis codes
- Documentation of failure of conservative treatment
- Duration of symptoms
- Relevant measurements and clinical findings
This isn’t about writing for the algorithm at the expense of patient care; it’s about ensuring clinical reasoning is captured in accessible, structured data. Physicians and coders should collaborate to create checklists for the top five or six services that most frequently generate denials.
7. Set Clear Guardrails Before Using AI Tools to Respond to AI Denials
Many vendors now offer tools that use AI to read denial letters and generate appeal responses. While these tools can be time-savers when dealing with high volumes of denials, they also have pitfalls. Some may cite policy language that does not exist or assert clinical facts not supported by the chart, errors that can be costly.
Establish clear rules before deploying any AI appeal tool:
- Every appeal generated by AI must be reviewed and signed by a human
- Every clinical assertion must be supported by the medical record
- Every policy citation must be verified against the payer’s actual policy document
Additionally, track the success rate of AI-assisted appeals compared to those written by staff. If the tool does not lead to a higher rate of overturned denials, it is not providing real value. Finally, confirm that the vendor has a business associate agreement in place, as the tool will be accessing protected health information.
8. Prepare Your Prior Authorization Workflow for January 1, 2027
Under the CMS interoperability and prior authorization final rule, impacted payers must have their required application programming interfaces (APIs) operational by January 1, 2027. These APIs will allow practices to submit requests and receive determinations, including specific denial reasons, electronically.
While the requirement falls on the payers, only practices prepared to use these APIs will benefit. Now is the time to ask your electronic health record (EHR) and practice management vendors what they are building to support these new interfaces, what additional costs may be involved, and when these features will be rolled out in your version. Also, check with your top payers to determine what level of support will be available on day one. Practices that are still faxing prior authorization requests in February 2027 will find themselves at a disadvantage compared to those instantly receiving structured denial reasons and integrating them into their workflows.
Conclusion
The landscape of prior authorizations and denials is evolving rapidly, driven by advances in automation and regulatory responses. Practices that adapt by understanding algorithmic denials, leveraging new legal requirements, systematically collecting and analyzing data, and preparing for digital interoperability will be positioned not only to minimize lost revenue but also to advocate more effectively for their patients. By treating denials as a challenge to documentation rather than a brick wall, and by investing in both legal and technological readiness, your practice can turn the tide in the ongoing struggle against automated and opaque payer denials.
Source: 8 ways to fight back when a payer’s algorithm denies your claim