🚀 Mission View
A sharper perspective on this week's issue that matters at the intersection of health and AI.
Two developments underscore just how large the ambitions for AI infrastructure have become: Sam Altman’s call to build gigawatts of new computing capacity each week (yes, you read that right - he wants to create a factory that can produce a gigawatt of new AI infrastructure every week) and Nvidia’s $100 billion commitment to expand OpenAI’s infrastructure. These are not incremental moves. They are big, bold bets on AI becoming as central to the global economy as electricity or broadband.
But for mission-oriented health organizations (community nonprofits, advocacy groups, provider networks, and smaller firms) the gap between these sweeping visions and day-to-day reality is enormous. Most leaders in health are still in the early phases of adoption or not even in adoption mode. They are grappling with skepticism, resource constraints, and questions about where AI can actually create value. We’ve seen this dynamic before with digital transformation in healthcare: the technology advances faster than the willingness or ability of people and institutions to adopt it.
That’s where the real leadership challenge lies. As Linda Hall, a Harvard Business School scholar of digital transformation, notes: the hardest part is never the technology. It’s the people. Big infrastructure bets may be necessary to unlock AI’s potential, but without leaders practicing productive urgency, urgency that is strategic rather than frantic, adoption in health will lag. The key for health organizations is not to mirror Silicon Valley’s pace, but to discern where AI can be integrated smartly, where it won’t work, and how to bring staff and stakeholders along in ways that build trust and impact.
Bottom Line: For health leaders, regardless of the billion-dollar bets being placed by tech titans, your competitive edge will come from leading people through thoughtful, step-by-step adoption that builds confidence and creates measurable impact.
🛜 Field Signals
A quick hit on this week’s key policy shifts and industry trends.
Google DeepMind released its Frontier Safety Framework 3.0, expanding AI risk monitoring to cover behaviors like “shutdown resistance” and “persuasive ability”. The update comes amid rising concerns about both large-scale societal threats and individual risks when users may be influenced to harm themselves or others.
Health insurers are increasingly using AI as part of their utilization management or prior authorization practices — a practice CMS is now eyeing for Medicare as well — raising concern about automated decision-making in health coverage. Congress may step in an to block the agency’s plans. Democrats and Republicans joined together in the House Appropriations Committee earlier this month to adopt an amendment that prohibits CMS from moving forward. Stil, as this shift unfolds, organizations may need to prepare for new regulatory scrutiny and compliance risks at the intersection of AI, health policy, and patient access.
AI isn't replacing radiologists but instead augmenting their workflows, as hybrid human-AI teams consistently show higher diagnostic accuracy than either humans or AI alone.
🛠️ Practical Edge
Actionable tips and tools to help leaders strengthen capacity and apply AI in their work.
STAT News has a useful rundown of six key groups now shaping standards for health AI, from newer collaboratives to long-standing trade associations. These organizations are helping health systems and developers decide how to evaluate, implement, and monitor AI tools responsibly - particularly as state and federal leaders debate the future of AI. For leaders, it’s a quick guide on who’s driving the rules of the road.
Researchers recently found that ChatGPT Plus often produces misleading summaries of scientific papers, favoring readability over accuracy. The (obvious) takeaway: Use AI as a first-pass assistant for drafting or rephrasing, but always fact-check and refine to ensure the science or information you are presenting stays correct.
Google launched Search Live, enabling real-time visual searches through smartphone cameras that combine Google Lens and Gemini AI models for contextual, conversational responses.
Microsoft officially added Anthropic’s Claude into 365 Copilot, marking the company’s first expansion outside of OpenAI for model choice. As the good folks at the Nueron noted, “Microsoft's basically admitting no single AI model rules them all … different models excel at different tasks. By letting users choose, Microsoft's effectively creating an AI model marketplace inside Office.”
At a time when AI is forcing managers must learn not just to adopt new tools, but to rethink workflows and lead teams through change, leadership expert Elizabeth Lyle’s argues that the next generation of leaders will need to challenge the status quo with speed, flexibility, and trust.
🌅 On the Horizon
A quick look at the developments and events expected to shape the weeks ahead.
Sep. 24-Oct. 15: The National Quality Forum (NQF), an affiliate of Joint Commission, is seeking comments from the public on a draft of its upcoming report on the use of artificial intelligence (AI) methods in healthcare quality measurement.
Sep. 29: LinkedIn News is hosting a webinar on “Build AI Habits That Set You Apart”, with Connor Grenna - CEO, AI Mindset and Chief AI Architect, NYU Stern School of Business. Register here.
Oct. 16: Crowell & Moring is hosting a webinar on the fast-evolving AI regulatory landscape, from the White House’s “America’s AI Action Plan” to a wave of new state laws in healthcare, labor, and privacy. The session will unpack compliance risks, federal–state tensions, and what organizations should expect next. Register here.
Oct. 21-22: TedAI San Francisco.
Dec. 8-9: Fortune Brainstorm AI San Francisco. Apply to attend here.
⌚️ Closing Time
A parting thought on what health leaders need to be focused on.
There’s growing concern about whether AI is living up to its promise. But as Fortune’s Eye on AI notes, the real issue may be a “learning gap” — organizations don’t yet know how to use the tools effectively, leading to low-value output or “workslop” rather than impact. Recent polling reinforces this: the 2025 AI Governance Survey found only 36% of small companies have someone dedicated to AI governance, 41% provide annual training, and roughly a third cite budget limits or lack of internal know-how as their main barriers. So what? For mission-oriented health organizations, the risk isn’t falling behind the latest model releases; it’s failing to build the people, skills, and governance structures that make AI actually useful.
Till next time,
