Google DeepMind and Google Research are expanding AI in medicine with tools and collaborations that span diagnostics, drug discovery, and public health forecasting.

The article emphasizes open-access, clinically oriented AI models and is edited by Diego Almada Lopez.

An AI system for mammography involving 125,000 women identified 25% of cancers missed during routine interval screenings and reduced specialist workload by up to 40%.

AlphaFold-assisted screening of 30,000 compounds in drug discovery contributed to the development of two FDA-approved drugs for Chagas disease.

Google’s flu forecasting model ranked first among 39 systems evaluated by the CDC for predicting hospital admissions for the 2025–2026 flu season.

MedGemma, an open medical AI model detailed in Nature Medicine, lowered barriers for AI-assisted molecular discovery and diagnostics, improving chest X-ray classification accuracy by 15.5 to 18.1 percentage points over prior models.

Google has committed over $60 million in funding for 2026, focusing on open-access models with clinical applications.

Google launched Antigavity, an interactive AI workbench to accelerate identification of antimicrobial peptides, leveraging ProtNLM and AlphaFold to prioritize candidates for wet-lab validation.

In partnership with Stanford, Google’s AI-enabled genetic diagnosis achieved a time-critical result in under eight hours, earning a Guinness World Record.

A collaboration with Imperial College London is part of the broader health AI initiative, though outputs were not detailed.

Overall, the announcements show a trajectory building on AlphaFold’s foundation to deliver practical health AI tools across diagnostics, therapeutics, and public health.