Sen. Bill Cassidy, the outgoing chair of the Senate health panel, released a discussion draft on 29 October 2026 that aims to promote access to health technologies that use artificial intelligence while setting guardrails, Bloomberg Government reported.
The draft bill is called the “Protecting Americans Through Intelligent, Ethical, and Transparent Artificial Intelligence in Health Care Act”, or the PATIENT AI Act. Its table of contents lists 19 sections after the title and definitions, split across four titles covering HHS, the Food and Drug Administration, the health IT office and Medicare and Medicaid.
Short explainer · Cassidy AI health bill draft
A vertical news Short narrated from this piece’s headline and opening line.
Business Pill · AI Agents
An evergreen Business Engineer explainer on AI agents. It explains the general idea only and does not discuss the Cassidy draft or health care.
The key insight: By our arithmetic, five of the PATIENT AI Act draft’s 19 sections, about 26%, sit in its Medicare and Medicaid title, which covers coverage frameworks, payment methodologies, billing codes and a CMMI payment model for applications that deliver care autonomously, according to the section-by-section summary.
In a statement quoted by Bloomberg Government, Cassidy said: “We’re living in a new digital age as artificial intelligence transforms how we deliver care.”
He added: “Implementing AI in health care must be done responsibly and to the benefit of patients, innovators, and consumers.”
Substantive sections of Cassidy’s PATIENT AI Act discussion draft by title: HHS general activities 6, FDA 5, health IT office 3, Medicare and Medicaid 5. Excludes the short-title and definitions sections. 5 / 19 = about 26%, our arithmetic. Source: discussion draft and section-by-section summary linked by Bloomberg Government, 29 Oct 2026.
Paying for AI Care
Title IV, on the Centers for Medicare & Medicaid Services, has five sections. According to the section-by-section summary, they direct the HHS Secretary to create a coverage evaluation framework for artificial intelligence-based services and a coverage evaluation pathway for applications reviewed by certified entities.
The summary says the Secretary would also develop proposed payment methodologies to reimburse for artificial intelligence-based services, and establish a process to use HCPCS Level II codes for them.
Section 405 would set up a payment model through the Center for Medicare and Medicaid Innovation on how to reimburse for use of artificial intelligence-based applications “that deliver care autonomously”, the summary says. By our arithmetic, these five payment and coverage sections are about 26% of the draft’s 19 sections.
FDA Review and Liability
Title II, on the FDA, has five sections. The summary says Section 202 establishes a risk-based approach to review software-based devices, including the ability for the FDA to accept premarket software changes. Section 203 would have HHS certify outside bodies to review low- to moderate-risk software.
Section 205 “reiterates existing federal preemption of state product liability laws for medical devices to apply to integrated and software-based devices”, according to the summary.
On liability more broadly, Section 102 directs HHS to enter into an agreement with a nonprofit organization to examine and develop recommendations on liability issues for artificial intelligence-based applications used in health care settings, the summary says.
Privacy, Transparency and Health IT
Under Title I, Section 101 directs HHS to issue guidance to HIPAA-covered entities on how to comply with privacy and security requirements when using artificial intelligence-based applications. Section 103 directs the Comptroller General to report on proposals for transparency and explainability measures.
Section 104 directs HHS to develop use cases for deploying artificial intelligence to improve internal operations and health care outcomes, which Bloomberg Government also highlighted. Under Title III, Section 303 would expand Certified Health Information Technology to include artificial intelligence-based applications not considered an FDA-regulated device.
The Structural Read
As we read the section-by-section summary, the draft pairs FDA review changes for software-based devices with Medicare coverage and payment work, so the same applications are addressed both as devices and as billable services.
As we read it, Section 205 restates existing federal preemption of state product liability laws for medical devices and applies it to integrated and software-based devices, while Section 102 sends wider liability questions to a nonprofit for recommendations.
By our arithmetic, the four titles hold 6, 5, 3 and 5 substantive sections, with the general HHS title the largest.
Sen. Bill Cassidy, statement quoted by Bloomberg Government
“Implementing AI in health care must be done responsibly and to the benefit of patients, innovators, and consumers.”
Three Implications
FOR AI HEALTH VENDORS The summary says Section 404 would establish a process to use HCPCS Level II codes for artificial intelligence-based services, and Section 403 would develop proposed payment methodologies to reimburse for them.
FOR SOFTWARE REVIEW According to the summary, Section 203 would have HHS certify bodies to conduct review of low- to moderate-risk software.
FOR PROVIDERS The summary says Section 101 directs HHS to issue HIPAA guidance on using artificial intelligence-based applications, and Section 105 directs training materials for health care entities.
The framework says: “SaaS is charged for access to tools. AGaaS charges for the execution of outcomes.”
As we read it, the draft’s Medicare sections ask how to pay for artificial intelligence-based services, including applications that deliver care autonomously. The framework was published by The Business Engineer and does not discuss the Cassidy draft or health care.
This is a discussion draft: the copy we read has a blank bill number and a blank sponsor line. Bloomberg Government’s report is paywalled beyond its opening, and we did not read a committee press release. Neither source we read gives a timetable for introducing or marking up the bill.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Cassidy’s PATIENT AI Act draft spreads 19 sections across HHS, the FDA, the health IT office and Medicare, and five of them deal with how Medicare and Medicaid would cover and pay for artificial intelligence-based services, including care delivered autonomously.
A note on sourcing. Facts come from Bloomberg Government’s report of 29 October 2026 (read to its paywall), the PATIENT AI Act discussion draft and its section-by-section summary, both linked by Bloomberg Government and read on 29 October. We did not read a committee press release or contact Sen. Cassidy’s office. Figures marked as ours are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
