1. FDA's Continuous, Adaptive AI Oversight Reaches a Key Milestone

Regulation / FDA / CMS

What happened: Most high-risk AI obligations under FDA's evolving framework take effect this month — a milestone building on last year's Predetermined Change Control Plan (PCCP) guidance, which replaced one-time device approval with continuous, adaptive oversight. January 2026 guidance simultaneously loosened premarket review for lower-risk tools while confirming that anything directly influencing clinical judgment still requires FDA review. 

Why it matters to our AIMed audience: The high-level turnover at the FDA has introduced much uncertainty. Regulation, however, is finally becoming what medicine has always been — continuous surveillance, not a single clearance. I think that health systems should mirror this internally: govern your AI projects the way you credential your physicians, with ongoing review rather than one-time privileges; perhaps someday we should certify teams/personnel rather than algorithms. 

 

2. China Deploys Medical AI at a Scale the West Hasn't Attempted

Global Health AI — Asia

What happened: China's ambitious 15th Five-Year Plan (2026–2030) calls for systematic application of AI across assisted diagnosis, precision medicine, health management, insurance services, and elderly care. The country has released roughly 300 medical large models, making AI core infrastructure for grassroots care. Chinese researchers themselves published a rare caution in JAMA on DeepSeek's tendency toward plausible but factually incorrect outputs. 

Why it matters to our AIMed audience: I am continuously impressed with China's unrelenting progress in not only AI but healthcare AI. Two governance philosophies are now running in parallel at scale: deploy-then-refine versus validate-then-deploy. The 68 million remote imaging cases thus far is mind-blowing. What the West should notice isn't the speed but the target: China is aiming AI for rural access, the use case US systems perpetually defer. 

 

3. AI Malpractice Litigation Gains Traction: Florida Delayed-Diagnosis Case

Governance & Legal Cases

What happened: A Florida lawsuit alleges AI contributed to a delayed cancer diagnosis — part of a growing wave of state-court litigation testing who is liable when AI contributes to patient harm. It joins insurer claims-denial suits against Cigna, UnitedHealth, and Humana, together forming the first real body of case law on algorithmic accountability in care delivery. There is also ongoing discussions about the non-use of AI for medical malpractice claims. 

Why it matters to AIMed audience: I have mentioned previously that medical liability exposure with AI for both use/non-use of AI is a board-level concern. We are watching malpractice law being rewritten in real time, and the medical legal issue of AI is moving faster than our governance committees. My concern is the emerging asymmetry: clinicians may soon face liability both for following and for ignoring AI. Careful documentation of the human-AI decision process is an available protection for now.

 

4. AI Biosecurity Risk Outpacing Governance, International Report Warns

General AI News — Biosecurity / Dual-Use Safety

What happened: The International AI Safety Report 2026, led by Yoshua Bengio and backed by 30+ countries, concluded AI capabilities in biological research are advancing faster than governance can keep pace. The UK is drafting safeguards against AI-enabled bioweapon design, OpenAI ran $50,000 bug-bounty challenges against its biosafety safeguards, and Anthropic published a framework urging government oversight of frontier catastrophic-risk testing.

Why it matters to AIMed audience: Biological risk sits squarely at the intersection of AI governance and public health security — a credibility marker for any comprehensive governance curriculum. What is new now is expediency: the same models accelerating our drug discovery compress the timeline for misuse. I think that health system governance committees should add biosecurity (as well as cybersecurity) literacy to their charters without delay.

 

5. Zuckerberg's "The Future Is for Everyone": Meta Pledges Superintelligence for All

General AI News — Big Tech / Strategy

What happened: Mark Zuckerberg published a 6,500-word manifesto titled "The Future is for Everyone," arguing the safest path to superintelligence is distributing it to many people rather than concentrating it in a few labs — resting on 3 principles: individual empowerment, invention as AI's primary purpose, and balance of power. He pledged everyone a personal AI agent, open-source biological models from Meta's Biohub, and a $1 billion community fund for data-center regions.

Why it matters: The manifesto directly challenges the safety-through-concentration vision championed by OpenAI and Anthropic — health systems must now navigate two competing governance philosophies. Both philosophies have merit — democratization and containment — but healthcare sits precisely where they collide. My prediction: "personal superintelligence" will reach our patients before it reaches any EHR upgrade ever will".