AI in Healthcare: Here Today, Gone Tomorrow?

AI is inescapable. It powers many of the ads we see on TV, shows up in our social media feeds, and is an essential tool for millions of working professionals. And in the exam room, it is rapidly becoming a permanent fixture for both clinicians and the patients they care for.

For decades, the role of AI in healthcare was mainly to explore potential applications for diagnostics and perform pattern recognition, among other research purposes. It influenced today’s technologies but remained distant from the exam room.

That distance has collapsed rapidly.

Today, AI is actively shaping how patients research symptoms, how clinicians review cases, and how care decisions are made. Yet despite its undeniable influence, the AI industry is more volatile than ever, with unresolved questions around legal liability and clinical reliability.

The tools clinicians rely on today could evolve dramatically—or disappear altogether—by tomorrow. So where do we go from here?

How AI made its way into the exam room

First, let’s take a brief look back at AI’s evolving role in healthcare over the years and how it arrived at its central role today.

AI’s earliest applications in medicine were academic and controlled. In the 1960s and 1970s, researchers experimented with rule‑based systems like Dendral or MYCIN, an AI system that assisted in the diagnosis of bacterial infections and recommended antibiotics. But despite showing remarkable promise, many of these systems were never widely implemented, partially due to a lack of funding and the extensive time it took to operate them.

By the early 2000s, the widespread adoption of electronic health records transformed healthcare’s data landscape, enabling a new wave of AI tools throughout the 2010s, particularly in imaging, risk prediction, and pattern recognition.

But 2022 marked a turning point.

When ChatGPT became accessible to the general public, the boundaries shifted overnight—placing it directly in the hands of providers and patients at the point of care.

Today, clinicians can access summarized histories, draft documentation, and AI-suggested next steps in seconds, just as patients can use the same tools to explore symptoms, gather information, and interpret their lab results. All with a single click.

When innovation meets oversight

These tools promise efficiency in a system already strained for time and resources—and in many cases, they deliver. Yet with major attention also comes scrutiny, and the sudden widespread use of AI has drawn the eyes of lawmakers across the country.

In the past few days, it was announced that a bill in New York proposes a ban on AI chatbots providing medical and legal advice. Several other states are also currently reviewing guardrails around AI’s role in healthcare, including proposals to establish standards for use and require safeguards around decision-making. Many have already taken action, from banning insurance companies from using AI as the sole decision-maker in prior authorization to legally requiring providers to disclose AI use to patients and obtain consent.

Headlines this past November further fueled concern when reports suggested that ChatGPT was restricted from providing medical advice to all users. Even if the reports weren’t fully accurate, the public uproar about this change revealed how tightly linked AI is to medicine today and how easily that access can be altered.

For the first time, we’re living in a moment where the tools both parties in medicine rely on could be reshaped or restricted with no warning.

Planning for the road ahead

There is no doubt that AI will always play a role in healthcare, but its long‑term usability in the exam room is far from settled. 

Scrutiny has now turned to how AI generates recommendations and when these can be relied on. Even when it works as intended, key questions remain:

- Who is responsible if the AI’s recommendation leads to an error?

- Is AI biased toward certain patient populations or data sets?

- How can AI account for unique cases or off-label prescription use not noted in guidelines?

As we look to the coming year, it’s possible that the systems we rely on today may not look the same a year, or even one month, from now. Capabilities could shift, and guardrails could tighten, forcing providers to adapt at the drop of a hat. 

In a landscape this fluid, one truth becomes even more important: clinicians need an “and.” Because while AI can accelerate data, synthesize information, and improve efficiency, it can’t stand alone, especially when its availability or scope is subject to forces outside the exam room. 

In the next blog, we’ll explore how patients and clinicians can navigate a world where health AI tools may be limited or removed, while still ensuring high-quality care is accessible.

Follow Healthcasts on LinkedIn, Facebook, and Instagram, then join our community to access insights from a community of verified peers today.