Google expands AI for medical query responses
Google announced on Monday (June 1, 2026) a significant update to Med-PaLM 2, its artificial intelligence system for medical diagnostics. According to Reuters, the new version of the model achieved 85% accuracy in controlled clinical trials, marking an important advance in the application of AI in healthcare.
Diagnostic accuracy
Med-PaLM 2 is based on PaLM (Pathways Language Model) technology and was specifically trained on medical data. In tests, the system demonstrated the ability to interpret symptoms, suggest differential diagnoses, and provide contextualized responses to clinical queries. The 85% accuracy rate was measured across a set of 500 clinical cases covering various specialties.
"The achieved accuracy places the system on a competitive level with general practitioners in controlled scenarios," said the Google Health research director in a statement.
Strategic partnerships
To validate practical use, Google has formed partnerships with hospitals in the United States and the United Kingdom. The institutions will test Med-PaLM 2 in patient triage, assisting in prioritizing cases and suggesting referrals. The goal is to reduce waiting times in emergency rooms and improve the efficiency of care flows.
In the US, the system will initially be implemented in three university hospitals. In the UK, the pilot project involves the National Health Service (NHS), focusing on emergency units.
Challenges and regulation
Despite the optimism, experts warn of ethical and regulatory challenges. Integrating AI into diagnostics requires continuous validation, human oversight, and protection of sensitive data. Google states that Med-PaLM 2 does not replace doctors but acts as a support tool.
Health authorities in the involved countries are closely monitoring developments. The FDA (US regulatory agency) has not yet approved the autonomous use of the system, but the current partnerships are considered validation tests.
Next steps
Google plans to expand access to Med-PaLM 2 to clinics and medical offices throughout 2027, subject to regulatory approvals. The company is also investing in multilingual versions adapted to different public health realities.
With this update, the tech giant reaffirms its commitment to innovation in healthcare, betting on AI as an ally for medical professionals in decision-making.