Google expands AI in healthcare: 6 Key announcements from ‘The Check Up’ event

AI enhancements in search

One of the major updates is an improvement in health-related search results. AI-powered Overviews, launched last year, are designed to provide more detailed and relevant responses to complex medical queries. Google claims users are now more satisfied with these results. The feature is also expanding to more countries and languages, including Spanish, Portuguese, and Japanese.

Additionally, a new feature called “What People Suggest” is being introduced to help users find information from personal experiences shared online. The company said this will allow individuals with specific conditions—such as arthritis—to easily discover advice on lifestyle adjustments from those in similar situations. The feature is currently rolling out on mobile devices in the US.

Medical records API in Health Connect

Managing personal health records is set to become easier with Google’s introduction of Medical Records APIs in Health Connect. The global rollout of these APIs will enable apps to securely read and write medical data, including allergies, medications, and immunisations, in a standard format, highlighted the American tech giant.

Google stresses that user data remains stored locally on devices, with individuals having full control over which apps can access it. The expansion means Health Connect now supports over 50 different data types, spanning activity, sleep, nutrition, and medical records.

Loss of pulse detection on Pixel Watch 3

Another development is the FDA-approved Loss of Pulse Detection feature on Google’s Pixel Watch 3. The function can identify when a user’s pulse stops due to a medical emergency—such as cardiac arrest or overdose—and automatically alert emergency services if the wearer is unresponsive.

Having already launched in 14 countries, Google says it will begin rolling out the feature in the US by the end of March.

AI in scientific research

Google is also introducing an AI system to assist biomedical researchers in generating hypotheses and research plans. Built on Gemini 2.0, the ‘AI co-scientist’ can process large volumes of scientific literature to propose testable theories.

The company is already collaborating with institutions such as Imperial College London and Stanford University, and it plans to expand access through a trusted tester programme.

AI for drug discovery

Alongside its research initiatives, Google has announced TxGemma, a collection of AI models intended to accelerate drug development. According to Google, these models are designed to understand and analyse therapeutic compounds, including proteins and small molecules, to help predict their safety and effectiveness.

Google says these models will be made available to researchers through its Health AI Developer Foundations later this month.

AI in paediatric cancer treatment

In partnership with the Princess Máxima Centre for paediatric oncology in the Netherlands, Google is backing an AI-powered tool named Capricorn. The system uses Gemini models to analyse medical data and patient records, generating treatment summaries for doctors.

By synthesising vast amounts of public and anonymised patient data, Capricorn aims to provide clinicians with comprehensive insights to inform treatment decisions, said Google.