Su Belagodu, Managing Partner at Intellectus Advisors, Fractional product and growth officer helping startups scale from idea to revenue.
getty
Healthcare has a reputation of being slow, cautious and allergic to change.
That reputation is not unfounded. This is an industry where electronic health record (EHR) migrations take years, and where a new clinical protocol can take a decade to move from research to bedside practice. The institutional logic is defensible; when decisions carry life-or-death weight, the cost of being wrong outweighs the cost of being slow.
But something is shifting.
A few years ago, the dominant conversation in healthcare AI was marked by resistance. Clinicians didn't trust it. Administrators didn't understand it. Compliance teams treated it like a liability waiting to happen. That conversation has changed materially.
According to the AMA's recent survey, 66% of physicians reported using AI in their practices in 2024, up from 38% in 2023, a 78% jump in a single year. Health system CIOs are no longer asking whether to implement AI. They are asking where to start. Boards that once flagged AI as a reputational risk are now flagging the absence of AI as a competitive risk.
Labor shortages, burnout, shrinking reimbursements and widening patient access gaps created conditions where standing still became more dangerous than moving forward. AI stopped being a future consideration and became an operational necessity, especially for the organizations with the least margin for inefficiency.
Healthcare has long operated on a short list of trusted technology vendors, a handful of established players whose names became synonymous with credibility. Breaking into that circle as a newer company was a challenge. That dynamic is not gone, but it is weakening.
The shift is partly generational. Clinical leaders who grew up with consumer technology carry different assumptions about what a credible product looks like. It is also structural. Large incumbents move slowly by necessity, their systems layered with complexity accumulated over decades. A health system with a specific, urgent problem cannot wait 18 months for a major vendor to scope a custom solution.
This has opened a lane for smaller, faster companies, but only for those willing to earn trust the hard way.
The companies gaining traction are winning by showing up differently, embedding clinical workflows, learning from real-world usage and iterating in ways that large vendors cannot.
Lexi, an AI-powered medical interpretation platform, learned this firsthand. In one of their first deployments, the team assumed that access would drive adoption. It did not. Staff kept reverting to phone interpreters, not because Lexi did not work, but because behavior in healthcare is deeply entrenched.
The team went on-site, observed real interactions and rebuilt the product integration around how staff actually worked. Siddharth Umarani Rajavelu and Linh Pham, Lexi's founders, described the goal as being focused on removing friction.
Abridge, a mature AI health company, does the same: They employ at-the-elbow training, where someone from the company is onsite and present during patient interaction. This is important because clinicians are not just learning a new tool. They are rewiring muscle memory built over years.
That arc, from skepticism to dependency, happens through proximity.
There is a harder conversation underneath the trust narrative, and it is worth having directly.
AI does restructure certain functions. Roles built around friction, coordinating, translating, transcribing and routing are changing. In a sector already navigating workforce sensitivity, that creates real organizational complexity.
The margin reality for community health centers reframes this. These organizations are not choosing between AI and a fully staffed team. They are often choosing between AI and a perpetually understaffed team that cannot meet patient demand. When AI handles the administrative and linguistic coordination that was previously bottlenecking care, clinical staff can focus on what only they can do.
Healthcare’s reputation for slowness has always been partly a story about distance. Large vendors controlled distribution. Procurement cycles favored established relationships. Innovation filtered in from the top of the market down.
AI is beginning to collapse that distance.
The organizations with the most acute needs and the least tolerance for slow, expensive solutions are now among the earliest real adopters.
What’s different this time is not just that the technology is transformative. It’s the proximity.
The pattern is becoming clear. Successful adoption is driven by how AI models are introduced into real workflows. Teams that are seeing traction are treating it as a capability to learn alongside.
In practice, that looks like:
• Working directly with clinicians during live workflows
• Training in real time rather than through documentation
• Iterating based on feedback, not assumptions
• Building trust before scaling usage
The companies driving real adoption are not relying on remote onboarding or self-serve models. They are showing up.
Healthcare is not becoming reckless about technology. The caution remains. But the definition of risk is changing. For a growing number of organizations, the greater risk is waiting too long to learn how to use it.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。