AGI and the future of medicine
For centuries, medicine has been a dialogue between patient and physician—a relationship built on trust, intuition, and the slow accumulation of clinical wisdom. That dialogue is about to be joined by a new voice. Artificial General Intelligence—systems capable of reasoning across any domain with human-level flexibility—does not merely offer doctors smarter diagnostic tools or faster data processing. It promises to fundamentally reimagine how we understand, treat, and prevent disease. A new kind of physician is emerging: one that never sleeps, never forgets, and can reason across the entire universe of medical knowledge.
The Autonomous Physician Arrives
The leap from narrow AI to AGI in medicine is not incremental—it is transformative. Where previous AI systems functioned as passive tools—flagging a suspicious scan, drafting a note, summarising a chart—agentic AI systems are becoming active participants in clinical workflows. They can analyse complex data, autonomously perform goal-directed tasks, and operate with growing independence.
The evidence is already here. In June 2026, researchers introduced MIRA (Medical Intelligence for Reasoning and Action), an autonomous AI agent operating in a sandboxed electronic health record environment. MIRA can navigate a large clinical action space to obtain patient histories; order and interpret laboratory, imaging and microbiology tests; generate differential diagnoses; and formulate treatment plans including prescribing medications and planning admissions. In simulations on real patient cases spanning multiple diagnoses, MIRA outperformed physicians in diagnostic accuracy and made guideline-concordant, medication-safe decisions.
The ambition extends further. The U.S. Advanced Research Projects Agency for Health (ARPA-H) is building ADVOCATE—the first FDA-authorized agentic AI capable of delivering around-the-clock cardiovascular care. The technology would be patient-facing, integrated in real time with electronic health records and wearable devices, and would act without routing every decision back to a clinician. The goal is software that could "do everything a clinician can over the phone". As the project lead framed it, the broader aim is a template "applicable to any patient with chronic disease".
Perhaps most remarkably, DeepRare—a multi-agent system for rare disease diagnosis—integrates more than 40 specialised tools and up-to-date knowledge sources. Across nine datasets spanning 14 medical specialties and 2,919 diseases, DeepRare achieved an average Recall@1 of 57.18%, outperforming the next best method by 23.79%. Expert review achieved 95.4% agreement on its reasoning chains. For the more than 300 million people worldwide affected by rare diseases—who often endure a "diagnostic odyssey" exceeding five years—this is not incremental improvement. It is a lifeline.
The Drug Discovery Revolution
The transformation extends beyond the clinic to the laboratory. In March 2026, researchers introduced Latent-Y, an AI agent that autonomously executes complete antibody design campaigns from text prompts—covering literature review, target analysis, candidate design, and selection of lab-ready sequences. Across nine targets, Latent-Y produced lab-confirmed nanobody binders against six, achieving binding affinities in the single-digit nanomolar range—without human filtering or intervention. Experts working with Latent-Y completed design campaigns 56 times faster than independent expert time estimates, compressing weeks of work into hours.
In June 2026, ATHENA-R1 was unveiled: an AI agent for treatment reasoning across all FDA-approved drugs since 1939, trained by reinforcement learning over a universe of 212 biomedical tools. Across five benchmarks of 3,168 drug reasoning tasks and 456 patient treatment cases, ATHENA-R1 reached 94.7% accuracy on open-ended drug reasoning and 82.9% on treatment reasoning—outperforming GPT-5 by 17.8 and 10.7 points respectively. In blinded evaluations by experts from 28 rare disease organisations, it was preferred over reference models on all criteria.
The pharmaceutical industry is taking notice. Insilico Medicine, a clinical-stage company specialising in generative AI for drug discovery, announced a strategic collaboration with Takeda in July 2026. The partnership aims to accelerate AI-driven drug discovery, with Insilico leading molecule identification while Takeda advances candidates through clinical validation. These are not laboratory curiosities—they are the new engines of pharmaceutical R&D.
Personalised Medicine at Scale
AGI is also realising the long-held promise of personalised medicine. By integrating multimodal data—electronic health records, genomic sequencing, wearable devices, and real-time biomarkers—AGI systems can tailor treatments to individual patients with unprecedented precision. For patients with diabetes, AI models can now predict changes in blood sugar levels and suggest optimal insulin doses, lowering the risk of hypoglycaemic events.
The convergence of AI with genomics and proteomics has revolutionised cancer research, enabling comprehensive analysis of complex molecular datasets. This integration provides a robust framework for linking molecular modifications in specific cancers, advancing precision oncology. As one analysis put it, precision medicine aims to promote multimodal data analysis using innovative AI approaches to understand disease mechanisms and discover significant biomarkers that could be used to prevent and predict complex diseases.
The Governance Imperative
Yet these advances come with profound governance challenges. Frontier AI systems—particularly agentic ones—represent something categorically different from traditional diagnostic algorithms. They raise at least four distinct governance issues: dynamism (risks are not static; systems update, drift, and develop emergent behaviours), autonomy (longer chains of actions with less human oversight), interaction (multiple AI systems coordinating can produce systemic risks), and context-dependence (risk emerges from the interaction between capabilities and deployment context).
These are not hypothetical concerns. A study published in NEJM AI reports that 43% of surveyed health systems are already piloting agentic AI, although only 3% have moved agents into live clinical workflows. Sixty-one percent of healthcare technology executives report building or implementing agentic AI initiatives or have secured budgets to do so. The question is no longer whether AGI will enter medicine—it is whether we will govern its entry with wisdom.
In February 2026, leaders from industry, government, policy, and academia gathered at New York University for a Summit on Building Governance Infrastructure for Frontier AI. The aim was ambitious and pressing: to develop governance principles before the technology outpaces the institutions responsible for overseeing it. As one ethicist observed, today's clinical AI systems are "teaching tomorrow's general intelligence what patterns constitute ethical behaviour". The ethical foundation we build now will shape the medicine of the future.
GFN's Role: Architecting the Medical Future
For Global Future Nexus, the transformation of medicine is inseparable from its mission at the convergence of AGI, planetary sustainability, and borderless human potential. GFN's Code of Ethics binds all members to principles ensuring trust, responsibility, and proactive stewardship across intelligences and systems. The President's Message frames this as the essential work of the organisation: advocating for and developing "frameworks for ethical emergence, legal identity ('Artificial Personhood'), and patient societal onboarding of AGI".
The AGI-Human Trust Building Labs—where humans and AGIs "live" each other's constraints—are essential laboratories for understanding how AGI can be deployed in healthcare without sacrificing accountability. A healthcare AGI that survives the "Triage Sandbox" understands that triage isn't math—it's trauma. The Fairness Committee's commitment to equitable access ensures that the benefits of AGI in medicine reach communities from Shanghai to Kigali, not just privileged centres. And the Governance Committee develops adaptive legal templates for city-state adoption—including the frameworks within which autonomous medical agents can operate.
A Future Worth Healing
The arrival of AGI in medicine is not an apocalypse. It is an inflection point. The question is not whether AGI will transform healthcare—it already is. The question is whether we will guide that transformation with wisdom, equity, and a deep commitment to the human relationships that make medicine meaningful.
As one physician-scientist put it, the true frontier is not technical capability but "responsibility". AGI in medicine is not about replacing doctors—it is about empowering them to do what they do best: heal. The physician of the future will not compete with machines. They will collaborate with them—and together, they will bring care to the billions who have never had it.
The diagnosis is clear. The treatment is being written now. The question is whether we will prescribe it with wisdom.
