How AI Is Changing the Face of Healthcare

AS ARTIFICIAL INTELLIGENCE (AI) goes mainstream, it’s making a big impact across the healthcare spectrum. In 2015, the U.S. Food and Drug Administration (FDA) approved just six AI devices (a glucose monitoring system, a snoring device, a retinal imaging system and a few others). In 2024, FDA approved 221 AI devices. And in just the first five months of 2025, FDA approved 147 devices for diagnostic and healthcare monitoring, ranging from improved MRI scans, cardiac amyloidosis and lymphoma.

One of the major differences in the healthcare industry regarding AI in 2025 is healthcare professionals’ comfort levels in using the technology. In fact, the risk tolerance for and confidence in AI technology has gone from low to moderate, and in some cases high (as long as safeguards are put in place).¹

Adam Rodman, assistant professor at Harvard Medical School and a physician at Beth Israel Deaconess Medical Center, expressed his optimism in The Harvard Gazette. “So what excites me most about AI and medicine?” he asked. “Well, the optimist in me hopes that AI and medicine can make us doctors better versions of ourselves to better care for our patients.”²

Alain Labrique, director for the Department of Digital Health and Innovation at the World Health Organization (WHO), offered this about the future of AI in medicine: “For us at WHO, AI is nothing short of a game-changer in public health, in clinical medicine and in maintaining our well-being as individuals.”³

And, Ashish Sukhadeve, founder and CEO of Analytics Insight, which provides organizations with strategic insights on disruptive technologies, made this bold statement recently in a Forbes article: “In terms of its transformative effect, AI is doing for healthcare what electricity did for the industry.”⁴

AI is indeed changing the face of healthcare in numerous ways. It is enabling personalized medicine and offering faster diagnostics and predictive analytics, as well as streamlining administrative workflow.

Enabling Personalized Medicine

AI is using the technology to tailor treatments to individuals by integrating data from genomics and wearable devices (smart watches, fitness trackers, etc.), which provide real-time data to monitor conditions such as heart disease and diabetes. AI combines behavioral and environmental factors to recommend personalized interventions that allow for customized, precise treatments to improve outcomes and reduce side effects.

A 2025 study explored how AI is transforming diabetes care by integrating real-time data from continuous glucose monitoring (CGM) devices with personalized treatment algorithms. In the study, type 2 diabetics who are not on insulin and who exhibit poor glycemic control used Dexcom CGM devices in combination with an AI platform called SugarFit Diabetes Reversal and Management Program that analyzed their glucose trends, lifestyle data and medication adherence. The 100-day study involved 1,752 patients (77.5 percent men; mean age, 50.22 years). Time in range increased from 45.74 to 49.31; time below range decreased from 7.46 to 5.34; and time above range decreased from 49.89 to 45.33. Patients also exhibited reductions in weight, HbA1c and fasting blood sugar.⁵

In a University of Southern California study, scientists used machine learning to analyze genomic and clinical data from more than 78,000 cancer patients across 20 different cancer types. Patients in the study received immunotherapies, chemotherapies and targeted therapies. The researchers discovered 95 genes significantly associated with survival in cancers such as breast, ovarian, skin and gastrointestinal cancers. Equipped with these insights, the team developed a predictive model to assess how patients with advanced lung cancer might respond to immunotherapy. The AI model helped clinicians match therapies to patients’ genetic profiles, enabling them to avoid ineffective treatments. The findings were validated using real-world data from collaborators at Genentech, Roche and Stanford University. Overall, the study demonstrates how AI can analyze large-scale genomic data to personalize cancer treatment and improve clinical outcomes.⁶

Faster Diagnostics

AI-powered tools are revolutionizing diagnostics in several ways. First, they’re analyzing medical imaging using deep learning (DL) models that detect anomalies in X-rays, MRIs and CT scans faster and often more accurately than human radiologists. Second, they’re processing electronic health records (EHRs), quickly sifting through vast amounts of patient data to identify patterns and suggest diagnoses. And third, they’re reducing diagnostic errors, providing decision support, helping healthcare workers avoid misdiagnoses and ensuring time-sensitive interventions. All these advancements are particularly impactful in areas such as oncology, cardiology and neurology, where early detection is critical.

A 2025 European Journal of Medical Research review examined the impact of AI across 16 diseases using machine learning (ML) and DL technologies. The study found that AI significantly reduced diagnostic time in key areas such as medical imaging (CT, MRI, X-rays), EHR analysis and predictive modeling for early disease detection. These tools not only improved diagnostic speed but also enhanced clinical decision-making and workflow efficiency, particularly in radiology and pathology. The authors concluded that ML and DL demonstrate “remarkable accuracy and efficiency in disease prediction and diagnosis,” ultimately strengthening patient outcomes.⁷

AI is also gaining time in administrative tasks. A 2025 study conducted at The Permanente Medical Group examined the use of ambient AI scribes across more than 2.5 million clinical interactions. Published in NEJM Catalyst, the study found that AI scribes saved more than 15,700 hours of documentation time in one year — the equivalent of 1,794 working days — while maintaining a neutral to positive impact on patient experience, indicating that AI did not detract from the human interaction during visits.⁸

Predictive Analytics

Predictive analytics powered by AI is helping healthcare systems anticipate patient needs, optimize resource allocation and prevent adverse events before they occur. Using advanced ML models such as DL neural networks, AI can analyze vast datasets, including EHRs, imaging and lab results to forecast disease progression, hospital readmissions or complications. For example, convolutional neural networks have shown high accuracy in predicting conditions like diabetic retinopathy and cancer spread, allowing clinicians to act earlier and more effectively.⁹

IBM Watson is processing vast amounts of medical journals and case studies. AstraZeneca’s AI is trained on data from half a million people, which allows it to predict diseases like Alzheimer’s before symptoms even appear.³

An AI algorithm called NAVOY Sepsis is reported to have the potential to predict sepsis hours before its onset.¹⁰ That means only a few hours after ICU admission, the clinical staff can receive high-performance risk assessment for sepsis in adult patients. The lifethreatening condition can ordinarily be difficult to detect because many people present with nonspecific symptoms, so time is always of the essence.

A scoping review performed in 2025 analyzed 142 studies applying AI to clinical risk prediction, including chronic diseases such as diabetes. It highlighted the use of ML, DL and causal ML for predicting disease onset and treatment outcomes. Some models achieved Area Under the Receiver Operating Characteristic curve scores up to 96 percent, indicating high predictive accuracy.¹¹

AI in Cancer Detection

One area in which predictive analytics is making a particularly strong impact is oncology. A new ChatGPT-style model called Clinical Histopathology Imaging Evaluation Foundation (CHIEF), designed by scientists at Harvard Medical School, trained on 15 million unlabeled images chunked into sections of interest. The tool was then trained on 60,000 whole-slide pathology images across 19 different types of cancer, including lung, breast, prostate and brain cancers. The model was trained to look at specific sections of an image, and the whole image allowed it to relate specific changes in one region to the overall context. CHIEF achieved nearly 94 percent accuracy in cancer detection.¹²

In one 2025 study involving 260,000 women, AI-assisted mammography increased breast cancer detection by 17.6 percent (95 percent confidence interval: +5.7 percent, +30.8 percent) higher than and statistically superior to the rate (5.7 per 1,000) achieved in the control group.¹³ The study compared outcomes between those screened with and without AI support. Additionally, the AI group had a higher positive predictive value for recalls than the control group.¹⁴

Streamlining Administrative Workflows

In addition to streamlining predictive analytics, AI is also transforming the administrative side of healthcare, where it’s helping to reduce paperwork, speed up processes and improve accuracy. Tools powered by AI are now automating tasks such as patient chart management, claims processing and billing and coding, allowing staff to focus more on patient care. For example, AI can automatically update EHRs, transcribe clinical notes in real time and even optimize nurse scheduling to reduce burnout and improve coverage.

A 2025 guide from Keragon, a healthcare automation platform, highlights how AI is being used to streamline operations across healthcare systems. From automating insurance workflows to analyzing data for compliance and resource planning, AI is helping healthcare organizations cut costs and improve efficiency. As these tools become more integrated into daily operations, they’re expected to play a key role going forward in reducing administrative burdens and enhancing the overall patient experience.¹⁵

Checks and Balances

As AI becomes more integrated into clinical settings, healthcare systems are implementing a range of safeguards to protect patients and ensure ethical use. Data privacy is a top priority, with regulations such as HIPAA requiring encryption, anonymization and strict oversight of patient data. Healthcare organizations are investing in cybersecurity and conducting regular audits to prevent breaches and misuse.

To combat algorithmic bias against marginalized populations, developers are being urged to use diverse and representative datasets. Bias audits and explainability requirements are helping ensure AI tools do not perpetuate healthcare disparities or misdiagnose marginalized populations.

Additionally, many AI systems are designed with a human-in-the-loop approach, meaning clinicians retain final decision-making authority, especially in high-stakes scenarios. On the legal front, countries are adopting varied approaches: Some, like the European Union and Japan, have introduced AI-specific laws, while others, including the U.S., are adapting existing frameworks. International collaboration is growing, with organizations like WHO promoting principle-based guidelines focused on fairness, transparency and accountability.

These safeguards aim to balance innovation with patient safety, ensuring AI enhances care without compromising ethics or equity.

A Look to the Future

As AI continues to evolve, its role in healthcare is expected to expand dramatically. Emerging technologies such as multimodal AI — which integrates text, images and sensor data — promise more holistic diagnostics, while federated learning allows AI models to train on decentralized data without compromising patient privacy. Innovations such as digital twins (virtual patient models) and AI-driven drug discovery are also on the horizon, offering new ways to simulate treatment outcomes and accelerate medical breakthroughs.

At the same time, healthcare systems are implementing critical safeguards to ensure AI is used ethically and responsibly. Together, these innovations and safeguards are shaping a future in which AI not only enhances care but does so in a way that is safe, equitable and patient-centered.

An Indispensable Partner in Medicine

As AI becomes more deeply embedded in the fabric of healthcare, its potential to improve outcomes, reduce costs and empower both clinicians and patients is increasingly evident. With thoughtful safeguards and ongoing innovation, AI is poised to become an indispensable partner in medicine, enhancing the capabilities of clinicians and empowering patients like never before.