AI Regulation Tracker  /  Regulation in force

Texas SB 815: AI Can No Longer Make a Medical-Necessity Denial, in Whole or in Part

Effective for Texas plans renewed on or after January 1, 2026, an automated system can no longer make a utilization-review denial, in whole or in part. The practical result is that a person has to make the medical-necessity call, and the state can audit the payer's AI. Here is how to use that when you appeal.

By Anthony Guerriero  ·  Reviewed by The Leveraged Years Editorial Desk  ·  Published June 27, 2026  ·  Last updated June 27, 2026

The denial that no one would own

Ask any physician who fights utilization review and you hear the same complaint. A medical-necessity denial arrives, the criteria are vague, and there is no person on the other end who will say they made the call. As payers fed those decisions through automated systems, the gap widened. Texas has now closed it.

SB 815, enacted in the 89th Legislature and effective for plans delivered, issued, or renewed on or after January 1, 2026, amends Insurance Code Chapter 4201. Per the enrolled text, a utilization review agent may not use an automated decision system to make, wholly or partly, an adverse determination. The statute defines an adverse determination as a finding that care is not medically necessary or appropriate, or is experimental or investigational. It defines an automated decision system broadly to include an algorithm that incorporates an artificial intelligence system. The practical translation for a treating clinician is direct. A denial cannot be the output of a machine. A person has to make it.

What the statute actually prohibits, and what it does not

The line SB 815 draws is between deciding and assisting. The prohibition is on using an automated decision system to make an adverse determination, in whole or in part. That phrase matters. It is not only AI-only denials that are barred; a denial that leans partly on an automated system to reach the no is also out of bounds.

What the statute preserves is the back office. Per the enrolled text, it does not prohibit the use of an algorithm, AI system, or automated decision system for administrative support or fraud-detection functions. So a payer can still use software to route, flag, and detect, but it cannot let that software conclude that a given patient's care is not medically necessary. Because the automated system cannot make the determination, the medical-necessity judgment has to come from a person, which is exactly who a treating physician needs on the other end of an appeal.

The audit power behind the rule

A prohibition is only as strong as its enforcement, and SB 815 gives Texas a real lever. Per the enrolled text, the commissioner of the Texas Department of Insurance may audit and inspect at any time a utilization review agent's use of an automated decision system for utilization review. That standing audit authority is the mechanism that lets the state look inside how a payer is using AI, rather than taking the payer's word for it.

For a clinician, that authority is a reason to document. A pattern of denials that look machine-generated, arrive faster than human review would allow, or carry boilerplate that does not engage the clinical facts is the kind of pattern a regulator can now examine. The audit power belongs to the department, not to the physician, but the physician's records are what give the department something to act on.

What the denial notice now has to tell you

SB 815 also tightens the paper. Per the amended Section 4201.303, a notice of adverse determination must include the principal reasons for the denial, the clinical basis for it, a description of and the source of the screening criteria and review procedures used as guidelines, and a description of the complaint and appeal process, including the right to an independent review organization. The change from earlier law is small but useful. The notice must now describe the criteria and procedures and name their source, not merely point at one or the other.

That gives the appeal a defined target. When the notice names the screening criteria, the clinician can argue the patient against that exact standard rather than against a moving one. When the notice names the source, the clinician can check whether the cited guideline even applies to the case. A denial that cannot produce a coherent criterion and source is a denial that is now harder to defend.

How to work a Texas denial under the new rule

The workflow is concrete. Read the notice first for the named screening criteria and the review-procedure source, because that is what the law now requires it to carry. Map your clinical facts to that exact criterion in the appeal, point by point, so the human reviewer has to engage the standard the payer chose. Note anything suggesting the determination was machine-made, including timing, generic language, or a failure to address the specific clinical record, and keep that note in the file. Then escalate with the independent review organization route the notice must describe.

The shift SB 815 makes is one of accountability. Before, a Texas physician could be arguing with a black box. Now the law bars the automated system from making the call in whole or in part, which leaves the medical-necessity determination with a person, the criteria behind it must be named, and the state can audit the payer's automated tools. The clinician who reads the notice closely and documents carefully is the one who turns that structure into a reversed denial.

Browse the full AI Regulation News tracker

Informational analysis for working professionals, not legal advice. Confirm how any rule applies to your situation with qualified counsel.