Tuesday, September 1, 2026

1. Signal of the Day · SNF / Post-Acute

   AI Is Now Drafting Nursing Home Plans of Correction — and the Pushback Has Started

   A post-acute compliance consultancy argues that generative AI can produce a Plan of Correction in minutes from a deficiency tag, but cannot perform the facility-specific root cause analysis a State Survey Agency actually evaluates — and that a POC is a binding commitment tested again at revisit. It marks the first widely visible case of skilled nursing operators putting generative AI on a document that creates legal obligation, arriving one week before the Risk-Based Survey goes national on September 8. Readers should note the publisher sells Plan of Correction services; the regulatory frame checks out against CMS State Operations Manual Chapter 7, while its claims about which AI tools exist and about regulators' own AI use are uncited.

   HighBy Brandie Elizaitis · CMS Compliance GroupRead at source ↗

2. SNF / Post-Acute Surfaced late · published August 21

   CMS Risk-Based Survey Goes National September 8 With Surveyor Materials Already Published

   CMS released the Risk-Based Survey surveyor package on August 13 inside the updated Long-Term Care Survey Process resources folder, including the RBS Entrance Conference Form, RBS CMS-802, RBS Pathways, Procedure Guide and Screening Job Aid, alongside a State Operations Manual Chapter 7 revision dated June 26, 2026. The operationally significant detail for qualifying facilities is that RBS entrance conference timeframes are shorter than the standard survey process, with all required information due by the end of the first day. Data-integrity audits of staffing and resident assessment submissions are among the disqualifying criteria, making survey eligibility partly an information-systems question.

   HighBy Brandie Elizaitis · CMS Compliance GroupRead at source ↗

3. AI Tech & Frameworks

   Frontier AI Labs Lean Into Health Care to Blunt Public Backlash Before IPOs

   Anthropic, OpenAI and peers are pushing into medicine and biology partly to reshape public sentiment as data-center opposition grows, with Nvidia and Eli Lilly building a $1 billion drug discovery lab and Isomorphic Labs raising $2.1 billion. Every named investment targets drug discovery, imaging and diagnostics — none targets long-term or post-acute care. For senior living and skilled nursing operators the practical implication is that AI will continue to arrive through EHR and staffing vendors governed by contract, not through the frontier research pipeline drawing the capital and the safety scrutiny.

   MedBy Adriel Bettelheim · AxiosRead at source ↗

4. Regulation & Policy Standing item · proposed rule published December 29, 2025

   HTI-5 Still Unfinalized, Leaving AI Model-Card Transparency as the Federal Baseline

   The ASTP/ONC proposal to cut 34 of 60 health IT certification criteria and strip AI model-card transparency from the decision support interventions criterion at section 170.315(b)(11) remains a proposed rule, with comments closed February 27, 2026 and no final rule published. Until it is finalized, HTI-1 model-card transparency continues to govern certified health IT, and it remains the principal federal mechanism by which provider organizations can see what AI their EHR vendor has embedded. This is the largest pending deregulatory instrument affecting health AI transparency.

   MedBy Healthcare DiveRead at source ↗

5. Regulation & Policy

   Executive Order 14421 Cites AI Growth in Declaring a Bulk-Power System Emergency

   Signed August 26 and published August 31 at 91 FR 55995, the order invokes IEEPA to prohibit acquisition or installation of foreign-produced bulk-power electric equipment tied to a Covered Foreign Entity, justified in part on the ground that the rapid growth of data centers and artificial intelligence has magnified national dependence on reliable electricity. It surfaces in AI sweeps but regulates grid hardware supply chains rather than AI systems, and imposes no obligation on healthcare providers. Energy Department implementing rules are due within 120 days and Federal Acquisition Regulation revision recommendations within 180.

   LowBy Federal RegisterRead at source ↗

Get the Daily Dispatch by email · free, about 8 a.m. Eastern · RSS