Inside 847 Production Clinical AI Notes — Sebastian Fox, Composo
Sebastian Fox from Composo analyzed clinical notes from AI tools that automatically write patient notes (known as ambient scribes).
He found that approximately one in twenty notes contained errors severe enough to harm patients, nearly one in five had important missing information, and more than one in ten contained fabricated information. Even when he built an advanced checker model to catch these errors, one-fifth of problematic notes still passed through. The biggest challenge is determining which differences between recordings and notes actually matter medically — a missing date may be insignificant, but missing symptoms such as jaw pain combined with new headache in patients over 50 could indicate a serious vision-threatening diagnosis.
Vibekollen prepared this summary with AI from the original publication. The content belongs to AI Engineer.
More to read
Lumaril, which develops AI software for optimizing industrial production by connecting to existing control systems, raised a $66M Series A at a $400M valuation
Techmeme 3 h ago
DeepSeek open sources Super-Agent as a research prototype, a proactive coding agent it says can monitor group chats for coding issues and suggest solutions
Techmeme 4 h ago
Insurance software maker Guidewire agrees to acquire Hyperexponential, which makes AI-powered insurance underwriting and pricing software, for $1.7B in cash
Techmeme 4 h ago
One word for AI agents in 2027
Nate Herk 8 h ago