Skip to content
Intelligence IAMay 1, 2026Intelligence IA
Article

Google DeepMind's 'AI co-clinician' outperformed GPT-5.4 in blind tests but still lags behind experienced doctors.

The system aims to assist physicians in patient care, but also shows why voice assistants like ChatGPT aren't ready for serious medical tasks.

Redaction Data Cube AISource: The Decoder
01

Brief source

Google DeepMind's 'AI co-clinician' outperformed GPT-5.4 in blind tests but still lags behind experienced doctors. The system aims to assist physicians in patient care, but also shows why voice assistants like ChatGPT aren't ready for serious medical tasks.