Soon, It May Be Unethical Not to Use AI in Healthcare

The future of healthcare is unevenly distributed, but rapidly arriving. To paraphrase futurist William Gibson: “The future of healthcare is already here; it’s just unevenly distributed.” Innovations like AI-driven diagnostics and predictive analytics already exist today – yet they aren’t accessible or utilized everywhere. This uneven distribution is closing fast as healthcare AI matures. And it raises a provocative idea: At some point, it will be unethical not to use AI in healthcare.When life-saving or cost-saving AI tools are available, choosing to ignore them will be hard to justify. The challenge now is preparing people and systems to embrace these tools, because change does not happen at the rate of technology – it happens at the speed of people. As one expert aptly put it, “Change does not happen at the rate of technology. It happens at the speed of people.” In healthcare, technology is racing ahead – and people must catch up.

The AI Revolution Outpacing Human Adoption

Healthcare has no shortage of AI innovations; what it lacks is widespread adoption. Over 1,000 AI-driven medical devices and algorithms have been cleared by the FDA to date – a number that has surged dramatically in just the past few years . Radiology alone accounts for roughly 80% of these approved healthcare AI applications , with algorithms now detecting cancers, hemorrhages, and fractures in medical images as accurately as top physicians . In fact, AI tools in radiology have become so advanced that failing to utilize them may soon fall below the standard of care. Yet despite this tech abundance, many clinicians remain slow to incorporate AI into practice.

Why the hesitation? The issue isn’t the algorithms – it’s our ability to integrate them. Healthcare culture is cautious and evidence-driven (for good reason), which means new tools face skepticism, workflow disruption, and regulatory hurdles. Adoption is lagging not due to a lack of AI solutions, but due to human factors like trust, training, and change management. As noted at NextMed Health’s AI workshop, there’s a “slow human adoption” curve even as technology sprints ahead. It’s a classic gap between what’s possible and what’s practical today. Bridging that gap requires showing clinicians and health leaders that AI is not here to replace their judgment, but to enhance it.

“Change does not happen at the rate of technology. It happens at the speed of people.” – NextMed Health AI Workshop (2024)

This quote resonates in healthcare: We can’t expect medical practice to transform overnight just because a new AI debuted. Change will happen when people trust and understand these tools. The good news is that momentum is building. Regulatory approvals for AI are accelerating (the FDA is now clearing ~20 algorithms per month) , and early adopters are demonstrating real value. For example, AI “scribes” that automatically write clinical notes have already become “table stakes” in many health systems , reducing burnout from documentation. Radiology groups are piloting dozens of AI models (from triaging critical findings to automating measurements) – and while full enterprise adoption is slow, the direction is clear. The future is arriving, and those who learn to use AI will set the new standard of care. Those who don’t? They risk falling behind ethically and economically.

Key Insights: AI’s Growing Impact in Healthcare

Let’s look at a few key insights that underscore why ignoring AI will soon be untenable:

• An Explosion of AI Tools: The U.S. FDA has now authorized over 1,000 AI-enabled medical devices and algorithms for marketing . Approvals have ballooned in recent years – from just 6 AI devices in 2015 to 221 in 2023 . These include AI for imaging, cardiology, oncology, dermatology, and more. Simply put, if there’s a repetitive cognitive task in healthcare, an AI solution likely exists or is in development. The tech is ready.

• Slow Adoption, Human Pace: Despite the tech boom, surveys show clinicians are proceeding cautiously. In medical imaging, only ~50% of organizations report using AI for at least one use case so far . And broadly, 60% of Americans say they’d be uncomfortable with their own provider relying on AI for diagnosis or treatment planning . This cautious mindset underscores that adoption lags innovation. As the saying goes, technology advances exponentially, but people change incrementally.

• Radiology Leading the Way: Radiology is a bellwether for AI’s potential. About 80% of all FDA-approved healthcare AI apps relate to medical imaging . AI can now detect certain abnormalities in scans with greater accuracy than expert radiologists , and it never tires – making it a perfect second reader. It’s telling that some radiology departments consider it malpractice to miss findings an AI would catch. AI in radiology is quickly moving from novelty to necessity, poised to become a standard part of reading images (much like PACS and digital imaging did).

• Massive Administrative Savings: Healthcare isn’t just clinical – it’s also bogged down by administrative tasks and paperwork. Here, AI’s impact could be revolutionary. Analysts estimate that automating routine administrative work (scheduling, billing, prior auth, data entry, etc.) could save **$150–$265 billion per year in the U.S. alone . Think of AI assistants handling insurance claims or managing care coordination behind the scenes. That’s hundreds of billions freed up for actual patient care. An often-cited stat: AI could cut ~45% of healthcare’s administrative activities – an efficiency leap that historically only comes once in a generation. Ethically, can we justify not streamlining operations when such waste undermines affordability and access?

• Data Overload vs. Intelligence Layer: Electronic health records contain a goldmine of data – but also a lot of junk. By some estimates, 30–50% of the data in current EHRs is duplicative, outdated, or irrelevant noise (copy-pasted notes, redundant labs, etc.). No human clinician can sift all this in real time, but an AI can. The vision is for AI to serve as an “on-the-ground intelligence layer” in care: continuously monitoring the flood of patient data, analyzing it, and highlighting the truly important bits to the care team. In other industries, AI is already the always-on assistant – filtering email, flagging fraud in finance, optimizing supply chains. In healthcare, an AI layer can flag that one abnormal result among thousands of vitals, or remind a doctor of a crucial detail about the patient’s history at the point of care . This kind of ambient intelligence can dramatically reduce errors and missed opportunities. It turns big data from a burden into a blessing.

Each of these insights points to the same conclusion: AI isn’t a futuristic nice-to-have; it’s rapidly becoming integral to quality healthcare. Whether it’s preventing a missed diagnosis, freeing clinicians from drudgery, or cutting costs that price patients out, AI will soon underpin the ethical, efficient practice of medicine. But for this to happen, we must address the valid concerns and fears that slow its adoption.

From Tech-Centric to Human-Centric: Rehumanizing Medicine with AI

It may sound ironic, but one of the strongest arguments for AI in healthcare is that it can make medicine more human. Today, many clinicians feel like data clerks – typing into the EHR late at night, wading through billing codes, struggling under paperwork – with less and less time for the human connection that is so vital to healing. Patients feel this too: ever shorter visits, eyes on screens instead of on them. AI offers a chance to reverse that trend, by automating the tasks that pull clinicians away from patients.

Imagine an AI assistant that transcribes and fills out your charts, orders routine tests, and even drafts up encounter notes – so the doctor can focus 100% attention on listening and empathizing during a visit. This is not far-fetched: these AI “scribes” are already being rolled out, and early studies show they can save physicians hours of documentation time each day . As renowned cardiologist Eric Topol said, the greatest gift AI can give doctors and patients is “the gift of time — to get back to… a deep bond with trust and empathy.” Free from onerous clerical work, healthcare providers can be fully present with patients again. In Topol’s words, AI can help restore the “humanity of medicine” this decade .

Of course, patients have understandable concerns about AI. Many worry it could depersonalize care or even make dangerous mistakes. A recent Pew survey found only 38% of Americans expect AI will improve health outcomes, while a third believe it could worsen outcomes . There’s fear of the unknown – will an algorithm decide my care? Will my doctor rely on a machine over their own judgment? These concerns must be met with transparency, education, and evidence. We need to clearly communicate that AI is a tool used by clinicians, not a replacement for them. It’s there to augment the doctor’s eyes and ears, not to silence their voice.

Notably, AI can actually rehumanize medicine when implemented thoughtfully. By taking over the rote tasks and crunching data in the background, AI lets humans do what only humans can – provide comfort, understand nuance, make ethical decisions, and build trust. A doctor freed by AI to spend 30 minutes with a complex patient (instead of 10 minutes plus 20 minutes of paperwork) can explore concerns, answer questions, and show empathy in ways that directly improve care. Meanwhile the AI might catch a subtle pattern in that patient’s records or vitals that helps guide a better treatment plan. It’s a powerful tag-team when done right.

Addressing the “black box” fear is also crucial. Clinicians and patients are more likely to trust AI if it’s not a mysterious oracle but rather a collaborative assistant that explains its reasoning. Newer AI systems strive to provide explanations for their suggestions (for instance, highlighting which X-ray regions triggered an alert for pneumonia). And rigorous validation in clinical trials is building confidence – many AI tools are now FDA-approved after proving they can perform on par with standard of care . As successes accumulate – an AI catching a cancer early that a doctor alone might have missed, or predicting a patient’s risk of complication in time to intervene – the comfort level with AI will grow.

There’s also the matter of AI’s well-publicized flaws, like “hallucinations.” Current generative AI, like ChatGPT, can sometimes produce confident answers that are completely incorrect. That’s scary if unchecked – nobody wants an AI making up a diagnosis or a fictitious medical guideline. The healthcare AI community is actively tackling this by refining models, limiting use to validated scenarios, and keeping a human in the loop for oversight. But there’s an interesting perspective emerging: these so-called hallucinations might also be seen as “creative leaps” the AI makes – going beyond the raw data into speculative territory. In a controlled setting, that creativity could even spark insights. The key is asking the right questions and prompts. As some AI experts quip, hallucinations are sometimes a user prompt problem. If you ask an ambiguous or overly broad question, you might get a dubious answer. The solution is not to throw out the AI, but to ask better questions. In other words, garbage in, garbage out – but a well-phrased query can yield brilliance. One AI commentator cheekily noted, “Hallucinations = creativity in disguise… Any bad output? Your prompt was the issue.”Instead of blaming the AI for its limitations, we should improve how we interact with it (and set appropriate boundaries for its use). With proper prompt engineering and domain constraints, the rate of AI errors drops dramatically . This is an important mindset shift: rather than viewing AI as an infallible oracle or a dangerous liar, we should see it as a highly advanced but fallible collaborator – one that we train, guide, and double-check, just as we would a human junior colleague. When used in this collaborative, monitored way, concerns can be managed while still reaping the huge benefits AI offers.

Alakin Health: Building an AI-Powered, Human-Centered Future

It’s one thing to talk about these ideas in theory – it’s another to build them into real-world solutions. Alakin Health is an example of a company embracing this vision of open, ethical, and powerful AI for good. Alakin has developed a no-code, AI-driven SaaS platform for care pathway design and patient engagement. In plain terms, it’s a cloud platform that lets healthcare teams easily create and automate digital care workflows – without writing a single line of code – and with AI intelligence woven throughout the process.

How does it work? Alakin provides a drag-and-drop care pathway creator that clinicians can use to map out a patient’s journey step by step . For example, imagine designing a 30-day post-surgery follow-up program: You could drag-and-drop modules for daily symptom check-ins via a chatbot, scheduled medication reminders, weekly telehealth video calls, and trigger rules for if the patient reports a concerning symptom. In traditional healthcare IT, setting this up would require custom software development or lots of IT support. With Alakin’s no-code interface, the doctors or nurses themselves can configure the pathway logic to their exact needs . This is incredibly empowering – it closes the gap between clinical intent and technical execution.

Once the care pathway is designed, Alakin instantly transforms it into a conversational patient app experience . The patient interacts through a mobile app (available in their language) that converses with them, collects data, and provides guidance. Here’s where the AI comes in: Alakin embeds AI at key points to personalize and optimize care. Its platform can do health forecasting – for instance, predicting which patients are at highest risk of complications or readmission, so the care team can proactively focus on them. It can automate patient engagement by using NLP (natural language processing) to understand patient-reported symptoms or questions and respond appropriately or alert clinicians if something needs attention. Essentially, Alakin’s system becomes an “on-the-ground intelligence layer” for the care team, monitoring all patients enrolled in a program and triaging information to the providers. Routine questions like “Is this side-effect normal?” can be answered instantly by an AI agent, while complex issues are flagged for a human nurse to follow up. This approach not only saves clinical time, it ensures no patient concern falls through the cracks.

Crucially, Alakin Health has a strong ethos of being open and ethical. The platform is built on interoperable standards like FHIR for health data , meaning it plays nicely with existing EHRs and data systems (no black box silos). Openness also means transparency – patients and providers can see the care protocols and how the AI is being used. Privacy and security are baked in (HIPAA compliant by design ). And by enabling providers to customize pathways or use trusted clinical protocols (e.g. leveraging standard care guidelines from ICHOM within the platform ), it ensures that the AI-driven automation is always rooted in accepted medical practice. Alakin’s team believes AI in healthcare must be “AI for good”– augmenting care, never compromising it. For example, any predictive algorithm they include is rigorously tested to avoid bias, and the ultimate decisions are left to humans; the AI offers guidance, not gospel. By focusing on improving the process of care delivery (making it more efficient, personalized, and responsive) rather than making isolated medical decisions, Alakin’s platform sidesteps many of the ethical minefields of AI in diagnosis. It’s using AI to empower clinicians and patients – not to push them aside.

Early deployments of Alakin show promising outcomes: higher patient engagement rates, reduced hospital readmissions, and much lower administrative burden for care coordinators. A care pathway that once required dozens of phone calls and manual data entry can now largely run itself in the background, with the care team intervening only when the AI flags a need. This is exactly the kind of real-world application of AI that makes one think, “How could we ever go back to doing this all by hand?” It’s analogous to using navigation apps in driving – you can drive without them, but having them usually means a faster, safer journey. In the near future, care designs and health management without AI assistance may seem just as inefficient and error-prone as navigating without a GPS.

By championing a no-code, human-friendly approach, Alakin Health also tackles the adoption barrier: clinicians and health admins can actually understand and control the AI-driven workflows. This demystifies the technology and builds trust. It’s a great example of building with AI and people together, rather than imposing AI on people. The company’s approach embodies the belief that AI in healthcare should be open (interoperable, transparent), ethical (responsible, patient-centric), and powerful (able to handle complex tasks at scale) – all aimed at the good of patients and providers.

A Call to Action: Build With AI, Not Against It

Healthcare innovators now face a pivotal choice. We can cling to the status quo, adding to the pile of burnout, ballooning costs, and unmet needs – or we can boldly embrace AI and shape its use for the better. History has shown that new technologies in medicine (from the stethoscope to the MRI) initially face resistance, but eventually they become indispensable standards of care. We are on that trajectory now with artificial intelligence. The difference is the sheer breadth of what AI touches – every specialty, every administrative process, every patient interaction could be elevated. That’s why the ethical imperative is emerging: if an AI tool can prevent an error or save a life, is it ethical to ignore it? If automation can save hundreds of billions and make healthcare more affordable, is it ethical to do things the old, wasteful way? If AI can give doctors back precious hours to truly listen to their patients, is it ethical to let them drown in paperwork instead?

The answer seems clear. The future is arriving rapidly, and it demands that we adapt. But adopting AI in healthcare doesn’t mean surrendering to a cold, robotic system. On the contrary, it means recommitting to what matters most – the human side of health – and using technology to support it. It means designing AI with empathy, applying it with responsibility, and always keeping the patient’s well-being and dignity at the center.

So, to healthcare leaders, clinicians, and innovators: let’s lean into this change. Educate yourselves and your teams about AI; pilot a project in your department; demand that AI vendors meet your ethical and clinical standards. Start small if needed, but start somewhere. The worst thing we can do is stick our heads in the sand while other industries leap forward. We have the chance to build with AI, not against it – to co-create a smarter healthcare system that works for everyone.

In a few years, we will likely look back and marvel at how we ever managed healthcare without the assistance of AI for routine tasks, information synthesis, and decision support. Just as we can’t imagine modern life without internet connectivity, the next generation of clinicians won’t imagine practicing without AI tools. And patients will not want to be cared for in systems that don’t leverage every available advantage to keep them safe and healthy.

The key is starting now, together, with intentionality. Let’s channel the current excitement into real improvements and share those success stories to build trust. By doing so, we usher in a new era of medicine that is not only smarter but also more human. The true promise of AI in healthcare is not that it will replace humans, but that it will elevate humans – enabling us to do what we do best, at a scale and efficiency never before possible.

The future is coming fast. Let’s make sure it’s a future where AI is saving lives and humanity is thriving – and where not using AI would be unthinkable.