AI Emerges as Healthcare’s Bright Spot Amid Industry Upheaval

Artificial intelligence is disruption for good, and powering healthcare innovation at a time when policy shifts and funding uncertainty are transforming traditional research and development end-to-end.

A new Cure survey released June 15, 2025, of healthcare entrepreneurs, investors and scientists found AI was a bright spot for these leaders as they navigate an ecosystem impacted by evolving regulatory frameworks, shifting funding models, and growing competition for talent. Their signal is clear: AI is not a sideline but central to their business models.

"AI and machine learning will transform the whole industry," one executive told Cure, a view echoed throughout The Future of Healthcare Innovation 2025, which reports the insights and strategies gleaned from the May 2025 survey of leaders and Cure interviews. More than half of the 116 respondents were CEOs or founders of private companies, and nearly 60 percent reported that the technologies are central to their organization’s business, investment or research focus.

That’s a huge shift for an enterprise that, for centuries, has been incremental by design,

building on trial and error to gradually accumulate life science knowledge. Many anticipate AI will upend the process, bringing efficiencies and opportunities even before lab studies begin, let alone clinical trial design, manufacturing, automating diagnostics and paperwork, and auto-piloting other brainy medical technology.

Investors are betting big on AI and digital health tools, according to CB Insights. Its venture backing for digital health companies increased 47 percent in the first quarter of 2025 compared to the last quarter of 2024, with $5.3 billion raised – the highest since the second quarter of 2022. Startups focused on AI accounted for 60 percent of that investment.

Unfolding the Promise of Proteins with AI

Irene Abrams, Senior Vice President for Research Innovation at the Boston Children’s Hospital Technology and Development Office, has worked with some of the most successful entrepreneurs and innovators in healthcare, including in recent years the developers of the blockbuster CRISPR drug Casgevy, a gene therapy for sickle cell anemia. She said AI research is gaining momentum, but will not yield results overnight.

“We all see the promise and the opportunity in AI, but the practical applications often lag behind,” Abrams told Cure in an interview.

One of the most promising applications so far, she said, is AlphaFold, a collaboration between Google and Isomorphic Labs. A tool made freely available to researchers, AlphaFold can predict protein structures without using physical, experimental techniques, such as X-ray crystallography.

“You can use it to predict the folding of every protein,” Abrams said.

A protein’s structure is critical information for researchers designing new drugs, because a protein’s structure reveals its function. And that tells researchers the role the protein plays in the life of the cell or organism to secure health or cause illness.

Simply by understanding how a protein folds, scientists can design drugs to enhance or inhibit the protein’s function and treat disease.

Cure Resident Companies Show Two Paths to AI-Powered Progress

Two companies in residence at Cure’s headquarters in New York City, have adopted seemingly opposite AI-enabled approaches to developing new treatments to meet unmet medical needs. Yet, both rely on AI to rapidly analyze vast amounts of biological information to match drugs to therapeutic targets.

Insilico Medicine, a clinical-stage biotech in which Cure’s affiliate Deerfield Management has a financial interest, uses generative AI and robotics platforms for drug discovery. Founder and CEO Alex Zhavoronkov, PhD, described the company’s platform as a tool that builds ranked matrices of disease pathways and therapeutic targets—allowing researchers to design and prioritize molecules with high commercial and clinical potential.

In the last four years, Zhavoronkov said,Insilico has advanced 10 drugs into early-stage clinical trials and identified 22 clinical candidates. In June 2025, Insilico reported progress on two candidates advancing as possible therapies, respectively, for a rare progressive lung disorder and for cancers. Rentosertib, a potential first-in-class molecule designed to treat patients with idiopathic pulmonary fibrosis, demonstrated safety, tolerability, and efficacy in a phase 2atrial. Also, an ENPP1 inhibitor showed positive data in treating hard-to-treat solid tumors in preclinical studies.

The second company is Every Cure, which, rather than designing new drugs, uses AI to match existing drugs with new indications, making it possible to speed therapies already proven to be safe and effective for other diseases to patients whose illnesses are currently untreatable.

The Every Cure story is proof of this approach. More than a decade ago, the company formed, not out of entrepreneurial zeal – at least not at first – but out of desperation. Co-founders David Fajgenbaum, MD, and Grant W. Mitchell, MD, then medical school roommates, launched their collaboration after Fajgenbaum developed Castleman’s Disease, a rare and fatal immune disease.

Fajgenbaum, Grant and other team members identified an overactive pathway in Fajgenbaum’s blood and realized that a 25-year-old generic treatment––the immune suppressant sirolamus used to combat kidney transplant rejection–could block it. Fajgenbaum has now been in remission for more than a decade.

To identify more FDA-approved drugs that could be used for off-label indications, the two co-founded Every Curein 2022, leveraging AI as a critical tool. In just three years, Every Cure reports that it has identified 14 of them, including arginine for sickle-cell disease, and a TNF-inhibitor, a drug used to treat inflammatory diseases, for a subtype of Castleman’s disease.

AI Could Lower Risk, Unlock New Therapy Options

Speaking with Cure about AI, Thomas Sakmar, MD, of Head of the Laboratory of Chemical Biology and Signal Transduction at The Rockefeller University, agreed that “evidence for optimal use of existing treatments” could be tremendously helpful.” He added that “AI-enabled passive surveillance and post-market follow-up for new therapies” might also provide valuable insights that can improve treatment outcomes.

Mitchell also noted that AI will likely expand the range of drug discovery, by reducing the financial risk for drug companies.

“Traditional pharma companies are risk averse and tend to develop products that are similar to historically successful medications,” he said. “AI and machine learning could potentially reduce the financial risks by developing digital models of disease states and potential therapeutics.

“If models are sufficiently accurate,” Mitchell said, “They will embolden researchers to venture into new areas without historical precedent.”

Read The Future of Healthcare Innovation 2025 for more opinions, concerns, and strategies of leading entrepreneurs, investors, academics, and government stakeholders.