專家警告醫療記錄中的人工智慧錯誤
Experts warn about AI errors in medical records
Updated at: September 7, 2026 at 01:45 AM
將人工智慧整合至醫療文檔中雖然有望提高效率，但同時也為患者安全與數據準確性帶來了重大風險。
The integration of artificial intelligence into healthcare documentation promises increased efficiency, yet it brings significant risks to patient safety and data accuracy.
專家們正對電子健康紀錄(EHR)中人工智慧的使用提出警示，並強調了幾個關鍵的危險。
Experts are raising alarms about the use of AI in electronic health records (EHR), highlighting several critical dangers.
其中一個主要擔憂是「幻覺」，即人工智慧生成看似合理但完全錯誤的資訊，例如偽造的症狀或並不存在的醫療病史。
One major concern is 'hallucination,' where AI generates plausible but entirely false information, such as fake symptoms or non-existent medical history.
此外，人工智慧系統經常難以處理轉錄錯誤，可能遺漏關鍵的臨床背景，進而導致危險的藥物衝突或錯誤的治療。
Additionally, AI systems often struggle with transcription errors, potentially omitting crucial clinical context that could lead to dangerous medication conflicts or incorrect treatments.
由於人工智慧識別的是模式而非理解醫療現實，它可能會放大分散式數據系統中已存在的錯誤。
Because AI identifies patterns rather than understanding medical reality, it can amplify existing errors found in fragmented data systems.
此外，關於固有的偏見以及臨床醫師所承擔的法律責任，也持續引發擔憂。
Furthermore, there are persistent concerns regarding inherent biases and the legal burden placed on clinicians.
即使在使用人工智慧時，醫師仍須對其紀錄負擔法律責任。
Even when utilizing AI, doctors remain legally responsible for their notes.
專家強調，人工智慧應發揮「副駕駛」的作用，而非作為自動化工具。
Experts stress that AI should function as a 'copilot' rather than an autonomous tool.
為了讓醫療保健從這些進步中受益，嚴格的人類監督至關重要。
For healthcare to benefit from these advancements, strict human oversight is essential.
臨床醫師必須仔細核對所有由人工智慧生成的文檔，以確保在日益自動化的醫療環境中，患者安全始終是首要任務。
Clinicians must meticulously verify all AI-generated documentation to ensure that patient safety remains the top priority in an increasingly automated medical environment.
