Scheduling note: This will be the last week we publish in 2025. We will resume our weekly distribution on January 9, 2026. Wishing everyone a happy and healthy holiday season.
🚀 Mission View: A sharper perspective on this week's top issues that matter at the intersection of health and AI.
Looking back across the year, a few patterns kept resurfacing. I’ve distilled them below into five core takeaways. The lesson isn’t that AI is overhyped or inevitable. It’s that how we choose to use it now will matter more than the tools themselves.
1️⃣ The hype cycle is real, but the ROI story is still forming
This year brought clearer evidence that heavy users of AI are seeing productivity gains. At the same time, enterprise adoption remains early, even as it feels like a new model drops from a frontier lab every week. Most organizations have not yet realized the workflow transformation often promised. The gap is not belief, but model reliability, workforce fluency, and redesigned ways of working.
Bottom line: Early productivity gains are real. Durable ROI will require better models, sustained training, and intentional workflow/organizational change.
2️⃣ The bottleneck is not just models. It’s organizational learning and trust.
Even as AI models improve, access to tools alone does not translate into effective use. Teams often lack shared understanding of how AI fits into real work, and they do not yet trust the systems, guardrails, or consequences enough to rely on them. Training without trust leads to workslop, shadow use, or quiet resistance.
Bottom line: Better models matter, but progress depends just as much on building confidence, fluency, and psychological safety across the organization.
3️⃣ AI is becoming a front door to healthcare, but it has limits
People have long used search tools like Google to look up symptoms, treatments, and questions before seeing a clinician. AI is extending that behavior by making information more conversational, personalized, and robust. Polling and user research suggest that while people are comfortable using AI to gather information or prepare for visits, they still want a human clinician involved for the most critical and sensitive decisions.
Bottom line: This may be a more powerful version of an old behavior rather than a wholesale shift in care. Whether AI ultimately complements or reshapes the clinician relationship remains to be seen.
4️⃣ Known and emerging AI risks are forcing the responsibility conversation
As concrete risks came into focus this year, from mental health chatbot harms to biosecurity and misuse concerns, the AI conversation shifted. Abstract ethics gave way to practical questions about accountability, intervention, and guardrails. That shift is where debates over human-in-the-loop, governance structures, and state versus federal roles gained urgency.
Bottom line: Responsible AI is not just a values discussion. It is an immediate and imperative operating requirement.
5️⃣ AI may either reinforce existing incentives or help reshape them
A central question this year was not whether AI works, but how it is being applied within today’s healthcare incentive structures. Layered onto existing systems, AI risks accelerating the very problems that have long frustrated patients, clinicians, and payers. Used differently, it could help shift priorities toward better outcomes and experiences.
Bottom line: Whether AI deepens existing dysfunctions or helps drive real improvement remains unsettled, and depends on how organizations choose to deploy it.
🛜 Field Signals: A quick hit on this week’s industry announcements, policy developments, and ethical considerations.
🏗️ Industry
AI adoption is rising, but strategy is lagging
New Gallup data shows employee AI use continues to climb, especially among knowledge workers. The gap is not awareness. It’s organizational clarity. Individual use is outpacing enterprise strategy, governance, and training.
Healthcare AI hype meets investor reality
Investors and founders are increasingly warning that AI enthusiasm in healthcare may be outrunning proven value. Expect more scrutiny of ROI, workflows, and durability as capital tightens.
Nvidia makes its open-model play
Nvidia’s release of Nemotron 3 is less about openness and more about ecosystem control, anchoring open innovation to its hardware stack.
AI agents are being used for real work, not errands
A Perplexity–Harvard study shows AI agents are primarily used for research, synthesis, and workflow management, not personal convenience tasks.
Audio joins the multimodal acceleration
Tools like SAM Audio from Meta signal how quickly AI capabilities are expanding beyond text and images, raising new questions about trust and verification.
🏛️ Government and Policy
Federal government turns to Big Tech for AI talent
The administration’s plan to bring technologists into short-term government roles reflects urgency around AI capacity and long-term questions about institutional expertise.
States step into the AI vacuum
Ohio lawmakers are advancing multiple AI bills, underscoring how states are still trying to fill federal policy gaps, even after President Trump issued his Executive Order last week attempt to preempt state laws.
a16z pitches Congress on a federal AI framework with room for states
Andreessen Horowitz is rolling out a proposed AI policy framework urging Congress to act on artificial intelligence, with an emphasis on child protections and preserving a meaningful role for states alongside any federal approach. The move signals how aggressively major tech investors are now trying to shape the regulatory terrain, not just respond to it.
Sanders calls for a data center moratorium tied to AI’s “unregulated sprint”
Sen. Bernie Sanders says he’ll push for a moratorium on construction of data centers powering rapid AI development, arguing it’s needed to let democratic oversight catch up and to ensure benefits aren’t captured only by “the 1%.” It’s a notable escalation: AI policy is no longer just about models and safety. It’s also about physical infrastructure, energy demand, permitting, and who gets to scale.
😇 Ethics and Responsible Use
A rare bipartisan agreement: wasteful care drives costs
Despite partisan gridlock, there is growing agreement that unnecessary care is a major cost driver, creating both opportunity and ethical risk for AI-enabled cost control.
🛠️ Practical Edge: Actionable tips, tools, and thoughts to help leaders strengthen capacity and apply AI in their work.
Zoom AI Companion 3.0 embeds AI into everyday collaboration
Zoom’s latest update integrates AI across meetings, chat, and tasks, making summaries and follow-ups default rather than optional.
OpenAI’s GPT-Image-1.5 makes image generation usable
Faster performance, better instruction following, and accurate text rendering finally make AI image tools practical for real workflows.
From SEO to AEO: preparing for the AI interface era
A strong deep dive on Answer Engine Optimization, AI agents as teammates, and the risks of low-quality AI content.
When leaders need to slow AI down
HBR outlines five tensions leaders must navigate in AI adoption, including the fast-versus-slow dilemma that derails many initiatives.
Note to my readers: I’d love to learn how you are using AI. If there’s a novel way you are deploying AI in your work, or seeing it utilized in healthcare, please feel free to shoot me a note and share: [email protected]
🌅 On the Horizon: A quick look at the developments and events expected to shape the weeks ahead.
👉 Mar 12–18, 2026 — SXSW 2026, Austin, TX
👉 Mar 30–31, 2026 — IAPP Global Privacy Summit, Washington DC
👉 Apr 6–9, 2026 — HumanX 2026, San Francisco, CA
👉 Apr. 10 — Ethical AI: Leadership and Governance (Virtual)
And finally, if you like what you are reading, please share this newsletter with your networks and encourage them to sign up.
Till next time,
BC
