About the Author
Josh Linkner is a five-time tech entrepreneur, New York Times bestselling author, keynote speaker, and globally recognized innovation expert. He has founded or co-founded five tech companies that sold for a combined value of over $200 million, and is co-founder and Managing Partner of Muditā Venture Partners, an early-stage venture capital firm. As Chairman of Platypus Labs, Josh helps organizations across industries build cultures of innovation and creative problem-solving.
In This Article
- Five macro shifts reshaping the industry, from AI-accelerated drug discovery and diagnostic imaging to the optimization of clinical trials and the rise of precision medicine
- Real examples of AI at work: Insilico Medicine’s AI-designed drug reaching Phase II clinical trials, the FDA’s surge past 1,000 approved AI radiology algorithms, AI-powered clinical trial recruitment at Cleveland Clinic, and more
- Three developments to watch over the next 12 to 36 months
- Why human creativity and judgment still matter more than any algorithm
- A 90-day action plan for healthcare and biotech leaders ready to move
In the healthcare industry, AI has moved well beyond pilot programs and research papers and into the daily operations of hospitals, pharmaceutical companies, biotech firms, and clinical research organizations. According to Grand View Research, the global AI-in-healthcare market was valued at approximately $36.67 billion in 2025 and is projected to reach $505.59 billion by 2033, growing at a compound annual growth rate of 38.9%. The AI-in-pharma and biotech market reached $6.63 billion in 2025 and is projected to hit $154.10 billion by 2034 at a CAGR of 43.55%. In 2024 alone, AI-driven biotech startups raised over $4.5 billion globally, with major recipients including Recursion Pharmaceuticals, XtalPi, and Insilico Medicine.
AI is rewriting the rules of how drugs are discovered, how diseases are diagnosed, how clinical trials are designed, and how treatment is personalized to the individual patient. It is compressing timelines that used to take a decade into months and opening therapeutic possibilities that conventional methods could never have reached. The winners in the healthcare and biotech space will be the organizations that move quickly and responsibly.
Here’s what you need to know.
Five Ways AI Is Changing Healthcare & Biotech
1. Drug Discovery Is Being Compressed from Years to Months
Developing a new drug has traditionally been one of the most expensive and time-consuming endeavors in any industry. Per-drug development costs can exceed $2.6 billion, and only about 1 in 9 drugs entering human trials ultimately gain approval. AI is fundamentally reshaping this equation.
Generative AI platforms can now design novel molecular structures, predict their biological activity, and optimize them for drug-like properties in a fraction of the time that traditional methods require. Insilico Medicine identified a novel target for idiopathic pulmonary fibrosis and advanced a drug candidate into preclinical trials in just 18 months, a process that typically takes four to six years, at a cost of approximately $150,000 excluding wet lab validation. Exscientia, in partnership with Sumitomo Dainippon Pharma, developed DSP-1181, the first AI-designed molecule to enter human clinical trials, in less than 12 months. Sanofi entered a $1.2 billion partnership with Exscientia to use AI to discover novel oncology and immunology therapies.
According to recent research, AI-developed drugs have demonstrated success rates of 80% to 90% in Phase I clinical trials, compared to 40% to 65% for traditionally discovered drugs. The implications for cost, speed, and patient access are enormous.
2. Medical Imaging and Diagnostics Are Being Transformed
Of all the areas where AI has gained traction in healthcare, medical imaging may be the most mature. AI algorithms can now detect patterns and anomalies in X-rays, MRIs, CT scans, and pathology slides that would be difficult or impossible for human clinicians to identify consistently at scale.
The regulatory numbers tell the story. According to a JAMA Network Open systematic review, 950 AI/ML devices had been authorized by the FDA as of mid-2024, of which 723 (76%) were radiology devices. By mid-2025, the FDA had added 115 new radiology AI algorithms to its approved list, bringing the total to approximately 873. By late 2025, the FDA’s public list of AI-enabled medical devices included 1,247 approved technologies, with more than 75% related to radiology. In 2025 alone, 295 AI/ML medical devices received FDA 510(k) clearance, with radiology accounting for over 70% of all clearances.
These tools are delivering measurable clinical impact. AI-assisted mammography has shown the potential to cut radiologists’ workload by nearly half without sacrificing diagnostic quality. AI triage of suspected strokes using CT scans has resulted in up to 30 minutes of saved treatment time. For healthcare leaders, AI in imaging has moved from experimental to operational, and the organizations that have not yet integrated these tools are falling behind.
3. Clinical Trials Are Being Redesigned from the Ground Up
Clinical trials face persistent challenges that AI is uniquely positioned to address. Eighty percent of clinical trials miss enrollment deadlines, pharmaceutical R&D costs exceed $200 billion annually, success rates sit below 12%, and data quality issues affect 50% of datasets.
AI is attacking these bottlenecks on multiple fronts. AI-powered patient recruitment tools have improved enrollment rates by 65%, identified eligible candidates three times faster by analyzing electronic health records, and achieved 93% accuracy in eligibility screening using natural language processing to read unstructured clinical notes. Predictive analytics models have achieved 85% accuracy in forecasting trial outcomes, and AI integration has accelerated trial timelines by 30 to 50% while reducing costs by up to 40%.
More than half of all clinical AI startups are now focused on patient recruitment and protocol optimization, according to the American Hospital Association’s analysis. The shift from static trial protocols to adaptive, AI-informed designs is changing what is possible in clinical research, and the organizations that embrace it will bring therapies to patients faster.
4. Precision Medicine Is Moving from Concept to Clinical Reality
For years, precision medicine has been discussed as the future of healthcare. AI is making it the present. By analyzing genomic data, proteomic profiles, electronic health records, and real-world evidence at scale, AI systems can identify which patients are most likely to respond to specific therapies, predict disease progression, and tailor treatment plans to the individual.
The drug discovery and development segment is expected to grow at a CAGR of 21.2%, the fastest of any application in the AI healthcare market, driven in large part by AI’s ability to match patients with targeted therapies. In January 2025, NVIDIA and Illumina announced a collaboration to integrate AI with genomics for enhanced multi-omic data analysis in drug discovery and clinical research. In May 2025, Danaher Corporation partnered with AstraZeneca to develop novel diagnostic tools intended to help clinicians determine which patients would most benefit from precision medicine treatments.
The convergence of AI, genomics, and real-world clinical data is creating a healthcare system that can treat patients as individuals rather than statistical averages. The organizations that build this capability will define the next era of care.
5. The Governance and Trust Gap Is Widening
Healthcare is among the most regulated industries on earth, and for good reason. The stakes of getting AI wrong in healthcare are measured in patient lives. This creates a fundamental challenge: the technology is advancing faster than the governance frameworks, workforce capabilities, and organizational cultures needed to deploy it safely.
The FDA issued its first comprehensive draft guidance on AI in drug development in January 2025, introducing a risk-based framework for model credibility that emphasizes transparency, validation, and data governance. In June 2025, the UK became the first country to join a new global network of health regulators focused on the safe and effective use of AI in healthcare. In January 2026, OpenAI acquired the healthcare startup Torch to integrate its medical memory technology into ChatGPT Health.
The organizations that will lead in this environment are the ones building governance infrastructure alongside their AI deployments, including data quality standards, algorithmic bias auditing, clinical validation protocols, and workforce training. The firms that treat governance as an afterthought will find themselves in a cycle of remediation and regulatory delay.
How AI Is Actually Being Used Today
Insilico Medicine’s AI-Designed Drug Reaches Phase II
Insilico Medicine is an example of end-to-end AI-driven drug discovery reaching clinical validation. The company’s proprietary platform, Pharma.AI, combines the biocomputational engine PandaOmics for target discovery with the generative chemistry engine Chemistry42 for molecular design. Using this platform, Insilico identified a novel target for idiopathic pulmonary fibrosis and designed a drug candidate, ISM001-055, advancing it to preclinical trials in under 18 months. In November 2024, the company announced positive topline results from its Phase IIa trial: patients receiving the highest dose showed a mean improvement of 98.4 mL in forced vital capacity from baseline, while the placebo group showed a mean decline of 62.3 mL. In June 2025, Insilico published the first proof-of-concept clinical validation of AI-driven drug discovery in Nature Medicine. The company now has over 30 drugs in development across cancer, fibrosis, immunity, and central nervous system diseases.
FDA-Approved AI in Radiology at Scale
The deployment of AI in medical imaging has reached an inflection point. By the end of 2025, the FDA’s list of AI-enabled medical devices surpassed 1,247 approved technologies, with radiology representing more than three-quarters of all authorizations. GE HealthCare leads with 115 radiology AI authorizations, followed by Siemens Healthineers at 86, Philips at 48, Canon at 41, United Imaging at 38, and Aidoc at 30. Tools like Viz.ai for stroke detection and MIT’s Mirai for breast cancer risk prediction represent AI systems that are already integrated into clinical workflows at major health systems. The shift from experimental pilots to production-grade diagnostic tools is happening now, and the pace of FDA authorization continues to accelerate.
AI-Powered Clinical Trial Recruitment
Clinical trial recruitment, long the single biggest bottleneck in drug development, is being transformed by AI-powered patient matching. At Cleveland Clinic, AI tools have accelerated patient recruitment by 170 times by scanning electronic health records to identify eligible candidates. Platforms like BEKHealth use AI-powered natural language processing to analyze structured and unstructured EHR data, identifying protocol-eligible patients three times faster with 93% accuracy. A comprehensive review published in ScienceDirect found that AI-powered recruitment tools improved enrollment rates by 65%, digital biomarkers enabled continuous monitoring with 90% sensitivity for adverse event detection, and predictive analytics achieved 85% accuracy in forecasting trial outcomes. For pharmaceutical companies spending billions on R&D, these are efficiency gains that fundamentally change the economics of drug development.
AlphaFold and the Revolution in Protein Structure Prediction
DeepMind’s AlphaFold has become one of the most consequential AI tools in biotech. The system predicts protein structures with near-experimental accuracy, and its predictions covering 98.5% of the human proteome have been made publicly available to the scientific community. Insilico Medicine used AlphaFold-predicted protein structures in combination with its generative chemistry platform to design a potential hepatocellular carcinoma drug in just 30 days, the first reported example of successfully using AlphaFold predictions to identify a confirmed hit for a novel target. Isomorphic Labs, a Google DeepMind spinout, announced a research collaboration with Johnson and Johnson in January 2026, leveraging a drug design engine that reportedly more than doubles AlphaFold 3’s accuracy on challenging protein-ligand structure prediction benchmarks. The practical utility of structural biology prediction for drug design is now established, and the implications for the speed and cost of discovery are profound.
What’s Coming Next: Three Moves to Watch (12-36 Months)
1. The First FDA Approval of an AI-Discovered Drug Is Within Reach
No AI-discovered drug has yet received FDA approval as of early 2026, though the first approval is estimated to have roughly a 60 percent probability of occurring in 2026 or 2027. The most advanced candidates include Insilico Medicine’s ISM001-055 for idiopathic pulmonary fibrosis and zasocitinib (TAK-279), now in Phase III trials through a partnership with Takeda. In December 2025, Takeda reported that zasocitinib eased the severity of plaque psoriasis in two late-stage trials. When the first AI-discovered drug receives approval, it will represent a watershed moment for the industry, validating the entire AI-driven discovery paradigm and accelerating investment across the sector.
2. Generative AI Will Reshape Hospital Operations and Clinical Workflows
Generative AI is expanding beyond drug discovery and diagnostics into the daily operations of hospitals and health systems. In January 2026, OpenAI acquired the healthcare startup Torch to integrate medical record aggregation into ChatGPT Health, signaling that large language models are being positioned as clinical workflow tools. Generative AI in healthcare is expected to grow from $2.65 billion in 2025 to $53.68 billion by 2035, at a CAGR of 35.10%. Applications include ambient clinical documentation (AI that listens to patient-physician conversations and generates notes), automated prior authorization, administrative workflow optimization, and real-time clinical decision support. The organizations that integrate these tools into their operations over the next 12 to 36 months will see measurable improvements in clinician productivity, administrative efficiency, and, ultimately, patient access.
3. AI Regulation Will Intensify and Become a Competitive Factor
The regulatory landscape for AI in healthcare is tightening on multiple fronts. The FDA’s January 2025 draft guidance introduced a risk-based credibility framework for AI models used in drug development, emphasizing validation, transparency, and data governance. The EU AI Act classifies most healthcare AI applications as high-risk, requiring formal explainability and bias auditing. The UK launched a new global network of health regulators focused on AI safety in June 2025. As of August 2024, the FDA had cleared 97% of AI-enabled devices via the 510(k) pathway, with regulators now actively exploring how to handle foundation models and large language models in clinical settings. Healthcare organizations that invest early in governance infrastructure, including algorithmic auditing, clinical validation, and regulatory documentation, will deploy new AI applications faster because the scaffolding is already in place.
The Human Factor: Why Creativity Still Wins
With all of this technological momentum, it is tempting to conclude that the future of healthcare belongs to whoever has the best algorithm. The reality is that the future belongs to whoever combines the best tools with the most creative, adaptable human thinking.
In my work with leaders across industries, I have seen a consistent pattern: the organizations that thrive in periods of disruption are the ones that cultivate what I call a Find A Way™ mindset, an organization-wide commitment to creative problem-solving that prioritizes agility over brute force and improvisation over perfect planning.
In healthcare and biotech, this matters enormously. AI can screen millions of molecular compounds for drug-like properties in hours, but it takes a human scientist to recognize the therapeutic insight that turns a promising compound into a viable treatment. AI can detect a subtle anomaly on a radiology scan with greater consistency than any individual clinician, but it takes a human physician to place that finding in the context of a patient’s life, symptoms, and values. AI can identify eligible clinical trial participants three times faster than manual review, but it takes a human researcher to design the trial that asks the right question in the first place.
The experience of AI drug discovery is instructive here. Despite the remarkable speed at which AI can generate drug candidates, several AI-designed compounds have missed endpoints in clinical trials for conditions including atopic dermatitis, schizophrenia, and cancer. These failures are a useful corrective: AI has not yet proven it can tame the stubborn failure rates of clinical medicine. The organizations that will lead are the ones that pair AI’s computational power with deep domain expertise and human judgment about what matters.
Start small and start now. Maybe it is your R&D team running a 90-day pilot on AI-assisted target identification for one therapeutic area. Maybe it is your imaging department benchmarking an AI triage tool against current workflows for a single scan type. Maybe it is a cross-functional workshop where clinicians, data scientists, and operations leaders sit together to identify the three AI use cases most likely to move the needle in the next 12 months. The breakthroughs accumulate, only if you start accumulating them.
Frequently Asked Questions
How is AI currently being used in healthcare and biotech?
AI is being deployed across the full healthcare value chain: AI-accelerated drug discovery and molecular design, medical imaging diagnostics (with over 1,247 FDA-authorized AI-enabled devices as of late 2025), clinical trial optimization and patient recruitment, precision medicine and genomic analysis, ambient clinical documentation, and administrative workflow automation. The global AI-in-healthcare market reached $36.67 billion in 2025 and is projected to exceed $505 billion by 2033.
How is AI changing drug discovery?
AI is compressing drug discovery timelines from years to months. Insilico Medicine advanced a novel drug candidate to preclinical trials in 18 months at a fraction of the traditional cost, and the compound has since shown positive Phase IIa results. Exscientia created the first AI-designed molecule to enter human trials in under 12 months. AI-developed drugs have demonstrated success rates of 80 to 90% in Phase I trials, compared to 40 to 65% for traditionally discovered drugs. The AI drug discovery sector drew $3.3 billion in venture funding in 2024.
How many AI medical devices has the FDA approved?
As of late 2025, the FDA’s public list includes 1,247 AI-enabled medical devices, with more than 75% related to radiology. In 2025 alone, 295 AI/ML devices received 510(k) clearance. The growth has been exponential: between 1995 and 2015, only 33 AI devices were approved, compared to 221 in 2023 alone. Radiology, cardiovascular, and neurology are the leading specialty areas for AI device authorization.
How is AI improving clinical trials?
AI addresses the most persistent bottlenecks in clinical research. AI-powered recruitment tools improve enrollment rates by 65% and identify eligible candidates three times faster. Predictive analytics achieve 85% accuracy in forecasting trial outcomes. AI integration accelerates timelines by 30 to 50% and reduces costs by up to 40%. Digital biomarkers enable continuous monitoring with 90% sensitivity for adverse event detection. At Cleveland Clinic, AI accelerated patient recruitment by 170 times.
What regulations are shaping AI in healthcare?
The FDA issued draft guidance on AI in drug development in January 2025, introducing a risk-based credibility framework emphasizing transparency and validation. The EU AI Act classifies most healthcare AI as high-risk, requiring formal explainability and bias auditing. The UK launched a global network of health regulators focused on AI safety. As of August 2024, the FDA had cleared 97% of AI-enabled medical devices via the 510(k) pathway. Healthcare leaders should treat governance and compliance infrastructure as a competitive investment.
Learn more about Josh Linkner’s keynote speaking in Healthcare & Biotech.
Josh Linkner speaks to healthcare organizations, biotech companies, and life science leaders around the world about innovation, navigating disruption, and building cultures that thrive in an era of rapid change. To explore how Josh can energize your next event, schedule a call today.