Washington Prior Authorization Transparency Act

Overview

Washington Senate Bill 5395, officially titled "Making improvements to transparency and accountability in the prior authorization determination process," represents a significant legislative effort to regulate the application of artificial intelligence (AI) within the healthcare insurance sector. Signed into law on March 23, 2026, and slated to take effect on June 11, 2026, this Act addresses growing concerns about the potential for AI systems to unduly influence or solely determine critical patient care decisions, particularly in the context of prior authorizations for medical services and prescription drugs. The legislation stems from a recognition that while AI can offer efficiencies in processing, human oversight by qualified medical professionals is indispensable when decisions impact patient access to necessary care.

The Act's foundational principle is to ensure that licensed health professionals retain ultimate responsibility for decisions concerning medical necessity, thereby preventing AI from becoming the exclusive arbiter of coverage denials. It introduces a comprehensive framework designed to inject greater transparency, accountability, and fairness into the prior authorization landscape. Key provisions include strict deadlines for health carriers to respond to authorization requests, a prohibition on the sole reliance on AI for denials, and enhanced disclosure requirements for the clinical criteria underpinning authorization decisions. By establishing these guardrails, Washington SB 5395 aims to safeguard patients from potential biases or errors in AI-driven determinations, reduce delays in accessing care, and foster a more equitable and understandable prior authorization process across the state's health plans.

Definitions

For the purposes of Washington SB 5395, several key terms are defined to ensure clarity and consistent application of the law. "Prior Authorization" refers to the process mandated by health insurance carriers, often termed "health plans," whereby approval must be secured before certain medical services, treatments, or prescription drugs are covered. This process is central to the bill's focus, as it is the point at which AI's influence is being regulated. A "Health Carrier" is broadly defined to encompass any entity offering a health plan, including traditional health insurance companies, health maintenance organizations (HMOs), and managed care organizations, all of which are subject to the Act's provisions regarding AI use.

The legislation also explicitly addresses "Artificial Intelligence (AI) in Prior Authorization," referring to the deployment of algorithms, machine learning models, or other AI technologies to assist in, or directly make, decisions concerning prior authorization requests. The Act's core intent is to prevent AI from being the sole basis for negative determinations, emphasizing the need for human intervention. "Medical Necessity" is another crucial concept, generally understood as healthcare services or supplies deemed reasonable and necessary for the diagnosis or treatment of an illness or injury, or to improve a patient's functional status. Decisions based on medical necessity are explicitly reserved for licensed health professionals under this Act. Furthermore, a "Licensed Health Professional" denotes an individual legally authorized to provide healthcare services, such as a physician, nurse, or other qualified medical practitioner, underscoring the requirement for clinical expertise in prior authorization denials. Finally, the Act touches upon "Retrospective Denial," which refers to a health carrier's denial of coverage for services after they have already been rendered, even if prior authorization was initially granted. The bill largely prohibits such denials, adding another layer of patient protection.

Governance and Institutional Framework

The governance framework established by Washington SB 5395 centers on reinforcing human oversight and accountability within the prior authorization process, particularly where AI technologies are employed. The Act explicitly mandates that licensed health professionals, rather than AI systems alone, must be responsible for making determinations to deny care based on medical necessity. This establishes a clear institutional hierarchy where clinical judgment holds precedence over algorithmic recommendations for adverse decisions. Health carriers are required to ensure that any AI tool used in the prior authorization workflow serves as an aid to human decision-making, not a replacement for it. This framework aims to mitigate risks associated with opaque AI systems and potential biases, ensuring that ethical and clinical considerations guide critical healthcare access decisions.

Furthermore, the Act places significant responsibility on health carriers to implement internal processes and structures that align with these new requirements. Carriers must disclose the credentials, board certifications, and specialty areas of the providers who had clinical oversight of any denial of prior authorization. This not only enhances transparency but also creates a clear line of accountability for adverse decisions. The legislation also strengthens the independent review process for enrollees disputing coverage decisions, allowing them to bypass certain internal grievance procedures when seeking external review. This institutional mechanism provides an additional layer of patient protection and external scrutiny, ensuring that carrier decisions, even those informed by AI, are subject to robust challenge and impartial evaluation. The overall framework is designed to integrate AI responsibly into healthcare administration while firmly anchoring decision-making authority in human expertise and ethical principles.

Key Focus Areas

Washington SB 5395 primarily focuses on several critical areas related to the use of AI in health insurance prior authorizations. A central focus is the restriction on AI as the sole determinant for denials. The Act explicitly prohibits private insurance companies and Washington's public employee health programs from using AI as the only means to deny, delay, or modify health care services for patients. This means that while AI can be utilized to process and approve requests, any denial based on medical necessity must undergo review by a licensed health professional. This provision directly addresses concerns about automated denials lacking human clinical judgment and potential for algorithmic bias to negatively impact patient care.

Another key focus area is enhancing transparency and accountability in the prior authorization process. The legislation mandates that health carriers provide clear and accessible prior authorization requirements, based on peer-reviewed clinical criteria that consider the needs of diverse populations. Carriers are also required to disclose the credentials of the providers overseeing any denial of prior authorization. Furthermore, the Act establishes strict timelines for prior authorization determinations, requiring responses within three calendar days for standard electronic requests and one calendar day for expedited requests. This aims to reduce delays in patient access to care. Finally, the bill includes provisions to prohibit retrospective denials for services that were previously authorized, with limited exceptions, and enhances reporting requirements for carriers to provide aggregated data on their prior authorization practices, including metrics on AI-aided denials and response times. These measures collectively aim to create a more equitable, efficient, and transparent system for patients and providers alike.

Implementation Framework

The implementation framework for Washington SB 5395 outlines a structured approach for health carriers to integrate the new regulations into their existing prior authorization processes. Upon its effective date of June 11, 2026, carriers will be required to immediately cease any practice where AI is the sole determinant for denying, delaying, or modifying prior authorization requests. This necessitates a review and potential overhaul of current AI-driven systems to ensure that a licensed health professional is always involved in the final decision-making process for adverse determinations. Carriers must establish clear internal protocols for human review, including defining the roles and responsibilities of the licensed health professionals who will provide clinical oversight. This involves training staff on the new requirements and updating operational procedures to reflect the mandated human-in-the-loop approach for denials.

Beyond the immediate changes to AI utilization, the implementation framework also includes requirements for enhanced transparency and reporting. Health carriers are tasked with making their prior authorization requirements more accessible and ensuring they are based on robust, peer-reviewed clinical criteria. This may involve updating public-facing documents, websites, and provider portals. Crucially, by October 1, 2026, carriers are mandated to begin reporting aggregated data on their prior authorization practices, including specific metrics on AI-aided denials and response times. This reporting obligation will require carriers to develop new data collection and analysis capabilities to accurately track and submit the required information to the relevant state authorities. The phased implementation, with the effective date preceding the reporting deadline, allows carriers a window to adapt their systems and processes before comprehensive data submission becomes mandatory.

Monitoring and Evaluation

The monitoring and evaluation mechanisms embedded within Washington SB 5395 are designed to ensure ongoing compliance and assess the effectiveness of the new regulations concerning AI in prior authorizations. A primary component of this framework is the mandatory reporting requirement for health carriers. By October 1, 2026, and on an ongoing basis thereafter, carriers are required to submit aggregated data on their prior authorization practices. This data must include specific metrics related to the use of AI, such as the number of prior authorization requests where AI was utilized, the number of denials influenced by AI, and the average response times for both standard and expedited requests. This regular data submission will provide state regulators with crucial insights into how AI tools are being deployed, their impact on denial rates, and whether the mandated timelines are being met.

In addition to quantitative data, the Act's emphasis on human oversight and transparency will likely be subject to qualitative evaluation. Regulators will monitor compliance with the requirement for licensed health professionals to oversee all AI-aided denials, potentially through audits of carrier processes and reviews of individual cases. The enhanced independent review process for enrollees disputing coverage decisions also serves as an indirect monitoring mechanism, as patterns of appeals related to AI-influenced denials could signal areas of concern. Furthermore, the legislative intent to increase transparency in prior authorization criteria will be evaluated by assessing the clarity and accessibility of information provided by carriers to both providers and enrollees. Through these multifaceted monitoring and evaluation strategies, the state aims to ensure that the Act's objectives of safeguarding patient care and promoting accountability in AI use are consistently met and that any unintended consequences can be identified and addressed.

Penalties, Liability, and Appeals

While the provided search results highlight the core prohibitions and requirements of Washington SB 5395, specific details regarding penalties for non-compliance are not explicitly detailed in the snippets. However, as a legislative act regulating health insurance carriers, it can be inferred that violations would typically fall under the enforcement powers of the Washington State Office of the Insurance Commissioner (OIC) or other relevant state agencies. These agencies generally have the authority to impose administrative fines, issue cease and desist orders, and take other disciplinary actions against carriers that fail to adhere to state insurance laws and regulations. The severity of penalties would likely depend on the nature and frequency of the violation, with repeated or egregious breaches potentially leading to more substantial sanctions.

Regarding liability, the Act's central tenet—that licensed health professionals, not AI, must be responsible for denying care based on medical necessity—shifts the ultimate liability for such denials back to human decision-makers and the health carriers employing them. This provision aims to prevent situations where responsibility could be diffused or attributed solely to an AI system, thereby ensuring clear accountability. The Act also strengthens patient appeal rights by enhancing the independent review process. Enrollees disputing coverage decisions are allowed to bypass certain internal grievance procedures when seeking external review, providing a more direct path for patients to challenge denials, including those influenced by AI. This mechanism for appeals is crucial for patient redress and serves as a vital check on carrier decisions, reinforcing the Act's commitment to patient protection and fair access to care.

Relationship to Other Instruments

Washington SB 5395 operates within the broader landscape of existing state and federal healthcare regulations, aiming to complement and strengthen patient protections. It builds upon existing prior authorization statutes by specifically addressing the emergent role of artificial intelligence. The Act is designed to integrate with established state laws governing health insurance carriers and the provision of healthcare services, such as those overseen by the Washington State Office of the Insurance Commissioner (OIC) and the Department of Health. By amending existing statutes related to prior authorization (e.g., RCW 48.43.830, 41.05.845, 48.43.525, 48.43.535), the bill ensures its provisions are legally integrated into the state's regulatory framework, rather than existing as a standalone, isolated measure.

While the Act is a state-level initiative, its principles resonate with broader national discussions and potential federal regulations concerning AI in healthcare and insurance. It aligns with a growing movement to ensure ethical AI deployment, particularly in sensitive sectors like healthcare where decisions directly impact human well-being. The legislation's emphasis on transparency, human oversight, and accountability for AI systems could serve as a model or influence future federal guidelines or legislation. Furthermore, by prohibiting retrospective denials for previously authorized care, the Act reinforces consumer protection principles that are often found in other state and federal insurance regulations. Its focus on reducing delays and improving access to care also aligns with the goals of various healthcare reform efforts aimed at streamlining administrative burdens and enhancing patient experience.

International Alignment

As a state-level legislative act in the United States, Washington SB 5395 primarily focuses on domestic regulatory concerns within Washington State. Therefore, direct international alignment is not its explicit objective. However, the principles underpinning this legislation reflect a growing global consensus regarding the responsible development and deployment of artificial intelligence, particularly in high-stakes sectors such as healthcare. The Act's emphasis on human oversight, transparency, and accountability for AI systems in prior authorization decisions mirrors ethical AI guidelines and proposed regulations emerging from international bodies and other jurisdictions worldwide. Many countries and international organizations, including the European Union with its AI Act, are developing frameworks that prioritize human agency, safety, and non-discrimination when AI is used in critical applications.

The concerns addressed by SB 5395—namely, the potential for AI to introduce bias, lack transparency, or make decisions without adequate human review—are universal challenges in AI governance. By requiring licensed health professionals to review AI-aided denials and mandating increased transparency in algorithmic decision-making, Washington State is contributing to a global dialogue on how to best integrate AI while protecting fundamental rights and ensuring public trust. While there are no formal treaties or international agreements directly referenced by this state bill, its approach aligns conceptually with broader efforts to establish ethical guardrails for AI, ensuring that technological advancements serve human well-being and do not undermine established professional standards or patient rights. This makes it a relevant example within the broader international landscape of AI regulation, demonstrating a practical application of ethical AI principles in a specific, critical sector.

Implementation Timeline

Compliance Checklist

Sources and References