The AI Healthcare Race Heats Up - But Are We Moving Too Fast?

Welcome back. There’s been no shortage of movement in healthcare and AI this week, and it feels like the race to reshape the system is officially on. What’s interesting is watching how private companies, governments, and investors are jockeying for position. Each with slightly different priorities, but all pointing toward the same future.

I came across Health Call’s acquisition by Northumbria Healthcare FT. On the surface, it looks like a classic scale-up play, but what stood out is the trust angle. Northumbria Healthcare clearly sees digital health as core to its future care delivery. Health Call has flown under the radar, but this move signals their ambitions are getting national, fast. It made me think about how few digital health companies actually get the nod from the NHS itself. That seal of approval could open doors others won’t walk through.

Speaking of strategy, I found it refreshing to read health tech leaders setting their digital priorities for 2025. What struck me was how aligned they are around getting the basics right. Data quality, interoperability, and making tech work for patients and clinicians. There’s nothing flashy about it, but maybe that’s the point. We’ve seen enough hype cycles. Now it’s about execution.

Meanwhile, the AI drumbeat is only getting louder. Google Cloud, no less, is pushing Agentic AI as the next frontier for improving patient outcomes. Essentially, AI agents that can act (not just analyse) within healthcare systems. The idea of an AI autonomously chasing test results or prompting care pathways is compelling. Still, I couldn’t help wondering where the human clinician fits if machines start "taking action" inside patient care. Are we setting up the next big ethical debate?

That blurred line gets even more pronounced when you read about NVIDIA and GE Healthcare’s new partnership on autonomous diagnostic imaging. The term "Physical AI" caught me off guard. It’s one thing to have AI models running in the cloud, but embedding that intelligence into imaging devices so they self-navigate and diagnose feels like we’re inching closer to AI taking over parts of medicine we thought were human-led forever.

If all this feels a little breathless, you’re not wrong. The World Economic Forum’s piece on AI transforming global health took a necessary step back. What stuck with me was the reminder that AI isn’t just about healthcare systems becoming more efficient, it could help tackle global shortages of doctors and infrastructure. Still, there’s a long road between potential and reality.

That gap between hype and execution was front and center in HT World’s feature asking how we move AI from hype to safe reality. The message was clear: governance needs to catch up fast. One thing I hadn’t considered was how “explainability” is going to be table stakes for AI in clinical settings. Doctors won’t just trust a black box, nor should they.

Then there’s the political side. If you missed it, NHS England is set to be scrapped and reimagined. The headlines are dramatic, but the underlying message is that the system isn’t fit for the digital age. Health tech leaders were quick to react, some welcomed the shake-up, others warned about unintended consequences. Either way, it’s a rare political moment where tech is seen as central to how healthcare gets restructured, not an afterthought.

And naturally, wherever AI moves, the money follows. The Wall Street Journal reported on VCs now seeing health AI as the next wave of returns. What caught my attention wasn’t just the flood of new capital, it was that investors seem more cautious, asking real questions about regulatory pathways and actual clinical impact. That’s new. And welcomed IMO.

Finally, none of this AI transformation happens without the grunt work of getting healthcare data ready for AI. This article was a good reality check. Before we race off building agentic AI or autonomous diagnostics, healthcare systems need to solve their messy data issues first. Otherwise, we’re building castles on sand.

Stepping back, it feels like this week’s theme is clear. Everyone sees AI as inevitable in healthcare, but the real work ahead is less about the technology itself and more about getting the foundations right. Governance, data, clinician trust, and making sure patients don’t get lost in the shuffle. There’s huge opportunity here, but as always, how we execute is what matters most.

Would love to hear your thoughts - especially on where you think the balance should sit between speed and safety as this next wave of health tech rolls out.

Thanks for reading.

Until next time!

Kevin McDonnell