We've all seen the headlines: 4 Ways Artificial Intelligence is Poised to Transform Medicine; Precision Medicine, AI, and the Future of Personalized Health Care; Can AI Transform the Way We Discover New Drugs? AI has taken the media by storm with extravagant claims and ambitious projections, but what are we actually seeing in daily practice, and what are the realistic applications physicians are looking forward to? Doximity conducted a survey across various specialties and career stages, asking which advancements in medicine physicians are excited about in 2025. Many cited AI, specifically personalized medicine and workflow applications, as something to look forward to this year. Here are a few ways we may see AI further shape health care in 2025.

Biologics Development

Since 2016, the Center for Biologics Evaluation and Research has seen an increase in AI and machine learning (ML) use in new drug applications for gene therapies. According to the FDA, there are currently more than 70 Investigational New Drug applications with AI/ML that involve “prediction, classification, clustering, and anomaly detection.” Gene therapies such as these are at the frontier of personalized medicine and have become a beacon for physicians and patients when traditional therapies fail.

More personalized treatment planning will require researchers and physicians to use AI at a more granular level to develop those more tailored biologics. 

Major pharmaceutical companies are already looking to use AI to help drug development. Amgen “trained a large language model on their proprietary data to help predict properties of proteins and develop biologics with enhanced properties,” while AstraZeneca is also interested in “digital biologics.” 

Albert Zhou, MD, a dermatologist in Connecticut, looks forward to seeing these new biologics in his practice. He anticipates a “greater armament of biologics and small-molecule inhibitors for psoriasis and atopic dermatitis.” He also says that this age of AI is an exciting time because “we can use AI to parse through a tremendous amount of information and distill down the information on many available medications [to find what works best for a patient.] Pharmaceutical companies are likely doing this as well, leveraging these tools, and AI may become an especially useful tool for diagnosis.” Dr. Zhou is hopeful for a future where “we have more targeted therapies to offer to patients with less collateral damage.”

Beyond existing use cases, Miguel Villagra, MD, an internist in Nevada, believes that algorithms for personalized treatment plans will gain traction this year. He expects the “integration of genomic data into everyday health care decisions will grow. Physicians leveraging telehealth could collaborate with labs and genetic counselors to offer precision treatment based on a patient’s genetic profile.” Dr. Villagra also envisioned that physicians could soon have the ability to “prescribe” apps or utilize wearable devices for continuous remote patient monitoring, aiding in the management of chronic diseases.

Workflow Applications 

In addition to targeted therapies tailored for patients, AI can be tailored to how a clinician works, with practical applications for searching through clinical references, acting as a scribe in a patient encounter, or quickly writing a prior authorization request, all features that Doximity’s AI Labs offer.

Many doctors anticipate that integration of data and algorithms will become part of daily workflows. 

"Generative AI will continue to grow in health care. Look for GPT tech to be embedded within EMR systems. [It] should help professionals comb through patient history much more efficiently,” said Vishal Gupta, DO, MPH, a neurologist in California. Dr. Gupta believes that with a few clicks, he will be able to see whether his patient has any contraindications with life altering treatments such as IV thrombolytics given during a code stroke, and facilitate decision-making in other high-stakes, critical care scenarios. “I could simply ask the AI, and I would have my answer in seconds,” he imagines.

The field of radiology, already transformed by AI, will likely see further advancements in 2025. In an article for STAT, radiologist George Lee Shih, MD, who also consults for OpenAI, said “I think that this kind of tool will be so powerful that essentially it’s going to read our minds at some point.”

As AI becomes more involved with health care, whether for decision-making, drug development, or other aspects, 2025, it seems, will not be the year AI changes medicine. It will be the year AI is medicine.

Illustration by Diana Connolly