Healthcare providers are increasingly deploying artificial intelligence (AI) tools for diagnostics, documentation and operational efficiency. In fact, over the last few months, large AI platforms are now marketing AI-enabled tools directly to healthcare providers. Providers must navigate a rapidly evolving regulatory landscape, while keeping in mind existing and longstanding requirements intended to safeguard the privacy and security of patient data. The US Department of Health and Human Services (HHS) has proposed updates to the HIPAA Privacy and Security rules, but without nationwide AI standards, existing HIPAA obligations continue to govern AI use. At the same time, providers may be subject to state regulation of AI. For instance, Texas providers face added complexity from new state-specific AI requirements. The central question emerges: How can providers remain compliant as federal and state rules continue to evolve?

HIPAA essentials

The HIPAA Privacy Rule (45 C.F.R. Part 164, Subpart E) governs the use and disclosure of protected health information (PHI). The HIPAA Security Rule imposes administrative, physical and technical safeguards for electronic PHI (ePHI) (45 C.F.R. Part 164, Subpart C). “Covered entities” include healthcare providers, health plans and healthcare clearinghouses. “Business associates” are persons or entities who create, receive, maintain or transmit PHI on behalf of covered entities.

Healthcare data privacy breaches often result from routine errors, such as misdirected electronic communications or unencrypted messages; external threats, including phishing, ransomware and hacking; and unauthorized access through vendor platforms. When a provider suspects such activity, it must investigate the source and mitigate any harm.

Providers face liability from the HHS Office for Civil Rights (OCR) and state attorneys general. Federal civil penalties are inflation-adjusted, with annual caps for repeated violations now exceeding US$2 million (see 2026 Annual Civil Monetary Penalties Inflation Adjustment). Criminal penalties for knowing violations can also include up to 10 years imprisonment.

Most states, including Texas, allow their attorneys general to enforce HIPAA violations through state consumer protection or health privacy statutes. State penalties can exceed US$250,000 per violation. Criminal violations can carry felony fines up to US$250,000 and more than 10 years of imprisonment, with even harsher penalties when identity‑theft or computer‑crime statutes apply.

Where HIPAA does not apply (for example, direct-to-consumer wellness apps), the Federal Trade Commission (FTC) oversees data security under the Health Breach Notification Rule (16 CFR Part 318), which covers both data security breaches and unauthorized disclosures.

The national and state AI landscape

HHS has not issued a final update to the HIPAA Privacy and Security rules addressing AI, but the federal government is moving toward a unified framework and leveraging existing health programs to shape expectations. On December 11, 2025, the Trump administration issued the Executive Order “Ensuring a National Policy Framework for Artificial Intelligence.” This order directs the DOJ to challenge onerous state AI laws and tasks agencies with creating preemptive reporting standards to avoid a "patchwork" of 50 different state rules. Additionally, on December 23, 2025 HHS issued a Request for Information on what HHS can do to accelerate the adoption and use of AI as part of clinical care.

HHS and the Centers for Medicare & Medicaid Services (CMS) have advanced AI through programs like the WISeR Model, a 2026 Center for Medicare and Medicaid Innovation model that uses AI to automate prior authorizations under traditional Medicare for a limited set of services. This model is now operating in six states, including Texas.

Despite federal efforts to harmonize AI, at least thirty-eight states have enacted AI-related legislation and almost 400 AI-related bills are currently pending in sate legislatures. Multi-state providers must track these varying obligations as there is no federal pre-emption for AI regulation.

Texas legislative changes

Texas has moved to the forefront of AI regulation. SB 1188 and HB 149 require healthcare providers to disclose to patients the use of AI in a clear and conspicuous manner.

SB 1188 (effective September 1, 2025)

Expressly permits AI for diagnostic purposes if the practitioner:

  1. Stays within the practitioner's licensed scope
  2. Discloses the use of AI
  3. Reviews all AI-generated records in accordance with Texas Medical Board standards

It also requires that electronic health records data be maintained within the US by January 2026.

HB 149, The Texas Responsible AI Governance Act (TRAIGA) (effective January 1, 2026)

Requires healthcare providers to disclose the use of AI in relation to healthcare services or treatment no later than the date of treatment. Failure to disclose may be considered a deceptive trade practice under Texas Law. Norton Rose Fulbright’s recent client alert further discusses TRAIGA’s requirements in detail, including for healthcare providers.

Where HIPAA meets AI

Although HIPAA predates modern AI, its requirements are intentionally technology‑neutral. The HIPAA Security Rule still applies, meaning covered entities must conduct an accurate and thorough risk analysis to identify vulnerabilities to the confidentiality, integrity and availability of PHI before adopting AI tools. Sharing PHI with an AI vendor will almost always require a Business Associate Agreement (BAA), and HIPAA remains the baseline standard for any AI system that handles PHI. Covered entities and business associates must maintain appropriate administrative, physical and technical safeguards and continue to perform the risk assessments emphasized by HHS for all ePHI.

In practice, this requirement means identifying AI use cases involving PHI, updating or executing BAAs with AI vendors, limiting access to approved tools and documenting risk analyses using recognized frameworks as federal and state regulations evolve. A “HIPAA‑first” governance approach, paired with close attention to emerging federal standards and state‑specific obligations, helps organizations stay aligned as the regulatory landscape shifts.

Practical recommendations (H.I.P.A.A. framework)

(H)onest disclosure

  • Clearly disclose AI use to patients, ensuring compliance with state-specific laws like TRAIGA in Texas.

(I)mplement safeguards

  • Establish policies, procedures and workforce training for approved AI tools.
  • Use robust data encryption and strict access controls.
  • De-identify data wherever possible before AI processing to minimize risk.
  • Consider retention settings associated with approved AI tools and whether this aligns with data minimization standards for maintaining PHI (see Section §164.502(b), describing the minimum necessary standard).
  • Ensure a licensed professional reviews AI outputs before sharing with a patient.

(P)HI sharing

  • Instruct users never to input PHI into public AI.
  • Execute BAAs with all AI software vendors.
  • Explicitly define permissible uses of data and prohibit the re-identification of de-identified datasets.

(A)udit AI use

  • Conduct thorough risk assessments for every tool, using resources like the National Institute of Standards and Technology (NIST) AI Risk Management Framework.
  • Maintain a written inventory of all technology assets that interact with electronic PHI.

(A)sk

  • The regulatory landscape is shifting rapidly. Our team of experienced lawyers and professionals at Norton Rose Fulbright monitors these developments closely to help you stay compliant.

The road ahead…

A thoughtful, well‑structured approach to AI governance is no longer optional for healthcare organizations in Texas and across other jurisdictions. As federal agencies refine national standards and states chart their own paths, providers that ground their AI strategies in HIPAA fundamentals, transparent patient communication and disciplined vendor oversight will be best positioned to adapt. By investing early in governance, documentation and cross‑functional coordination, organizations can stay ahead of regulatory change and deploy AI to strengthen patient trust and support long‑term operational resilience.

Contact