Google DeepMind announced its AI co-clinician research initiative on April 30, 2026, introducing an advanced system designed to complement human physicians through real-time video-based exams, diagnostic reasoning, and personalized treatment recommendations. The system demonstrated performance at or above primary care physician levels across multiple medical assessment areas.
AI Co-Clinician Performs at Primary Care Physician Level in 68 of 140 Areas
The AI co-clinician was tested using live audio and video to simulate telemedical calls, allowing it to observe physical cues including gait, breathing patterns, and visible skin changes. Performance results showed:
- Performed at or above primary care physician level in 68 of 140 assessed areas
- Won 67 to 26 in blind comparisons against an existing clinical AI system
- Outperformed GPT-5.4-thinking-with-search 63 to 30
- Achieved zero critical errors in 97 out of 98 primary care queries
- Expert physicians still outperformed the AI overall, particularly in identifying red flags and guiding critical physical examinations
Multimodal Capabilities Enable Physical Exam Guidance
The system's ability to process live video enabled new diagnostic capabilities. It could guide patients through parts of physical examinations, including checking inhaler technique and helping identify shoulder injuries. This multimodal approach represents a significant advance over text-only medical AI systems.
Triadic Care Model Positions AI as Augmentation Tool
Google DeepMind positions the co-clinician within a "triadic care" model where an AI agent works alongside both physician and patient rather than replacing clinical judgment. This approach emphasizes augmentation over automation, addressing concerns about AI replacing human doctors.
The research demonstrates Google DeepMind's continued investment in healthcare AI, combining advanced diagnostic reasoning with real-time multimodal analysis to create tools that support rather than supplant medical professionals.
Key Takeaways
- Google DeepMind's AI co-clinician performed at or above primary care physician level in 68 of 140 medical assessment areas
- The system outperformed GPT-5.4-thinking-with-search 63 to 30 in blind comparisons
- Live video capabilities allowed the AI to observe physical cues and guide patients through exam components like inhaler checks
- The system achieved zero critical errors in 97 of 98 primary care queries during extensive testing
- Google DeepMind positions this as a "triadic care" model designed to augment rather than replace physician judgment