Hospital AI Lab Practical information on medical AI for U.S. hospitals

Patient data, health records and AI — for administration and nursing

The decision to adopt AI in a hospital —
read from U.S. rules and primary sources.

Is the product a medical device, on what basis may patient data be used, where does it go, and how fast must a breach be reported. Collected from the eCFR, the Federal Register and FDA pages; each item gives its section and the date checked.

Roadmap New federal documents

Updated:

When you face thisBY PURPOSE

UpdatesWHAT'S NEW

See all →

Last 7 days: AI news 47 · New federal documents 0 · Training and events 0 · Videos 15

New federal documents from the Federal RegisterNEW REGULATIONS

See all →

The date is the publication date of the document, not a deadline. Read the official text for its content and effect; the list reflects documents of the selected agencies, not only those that concern hospitals.

Featured articlesARTICLES

All articles →

Read by topic

All series →

Our principlesOUR PRINCIPLES

  • 1
    Based on primary sources

    We write only what comes from official texts that were opened and read (eCFR, Federal Register, FDA, ASTP/ONC), with links.

  • 2
    The date checked is visible

    At the bottom of every practice and patient page: “Sources and date checked.” Rules change, so the date is required.

  • 3
    Source links are checked daily

    The English edition has 43 external links; they are checked automatically every day, and 0 are not found (404/410) (November 16, 2026).

  • 4
    No recommendations of particular products or hospitals

    We look at whether something can be used in your hospital and on what basis, not at brands or a push to adopt quickly.

How material is written and checked →

All pages at a glanceALL PAGES

Learning materials and referenceLIBRARY

Questions about this siteABOUT THIS SITE

Questions about medical AI →
Who is this site for?
Mainly people in hospital administration and nursing: planning and finance, IT, compliance, and nurse managers. It is not a site only for physicians. Deciding on adoption, reviewing a proposal and managing the budget are administrative work. Responding to system alerts and adjusting workflows is nursing work.
Where does the information come from?
AI news headlines are collected automatically from trade and national outlets (Fierce Healthcare, Healthcare Dive, Becker’s Hospital Review, STAT, MedCity News, HIT Consultant, Radiology Business, Health Affairs, KFF Health News, NEJM AI, JAMA Network, Healthcare Business Today, TechCrunch, The Verge, Ars Technica, WIRED, MIT Technology Review, The New York Times, NPR, BBC News and The Guardian), and new federal rules come from the Federal Register. Statements about regulations in the practice guides, glossary and checklists carry a citation to the primary text: the eCFR, the Federal Register, HHS, FDA, ASTP/ONC and CMS.
How often is the site updated?
News is checked several times a day, new federal documents several times a day, and the links to sources and the editions of the core regulations are checked daily. A page changes only when something new appears. Each practice guide shows the date it was last checked.
Is this a translation of the Korean edition?
No. The English edition is written from scratch for the United States, from sources such as the HIPAA rules in 45 CFR Parts 160 and 164, FDA guidance on AI-enabled device software, and the ONC rules in 45 CFR Part 170. The content differs from the Korean edition.

About this siteABOUT

Hospital AI Lab is a site for people who decide whether a hospital adopts medical AI. It is not a technical manual for developers. The pages cover the questions that administration and nursing run into: the regulatory status of a product, patient data, and day-to-day operation. The content is general information and does not replace advice from a lawyer, a compliance officer, or a qualified specialist for your situation.

The Korean edition

The main edition of the site is in Korean and has been running since 2026. The English edition is newer and has fewer pages.

Korean edition (한국어) →