返回新闻
产品展示AI Understanding 简报

OpenAI 通过 Epic 将 ChatGPT 连接到授权的患者记录

Scripps News 报道称,OpenAI 已宣布为 ChatGPT 医疗保健提供 Epic 集成和医疗保健公共数据插件,允许授权临床团队审查患者信息。

4 min readRead the linked source
Source-provided image accompanying OpenAI links ChatGPT to authorized patient records through Epic
来源参考来源记录
出版商
scrippsnews.com
来源链接
scrippsnews.comhttps://www.scrippsnews.com/science-and-tech/artificial-intelligence/openai-announces-new-tools-that-connect-chatgpt-with-authorized-patient-records
来源类型
链接来源——主要来源状态尚未确定。
背景60 秒内了解这一点

从这里开始

关键术语

特征
模型用来进行预测的输入变量。
测试一下自己什么是人工智能?测验

发生了什么

Scripps News reports that OpenAI announced tools connecting ChatGPT with authorized patient records through a new Epic integration and a Healthcare Public Data plugin. OpenAI says clinicians can use them to review patient histories, lab results, medication changes and follow-up needs before appointments. The report identifies two versions: ChatGPT for Clinicians and ChatGPT for Healthcare.

Scripps News says OpenAI announced a new Epic integration and a Healthcare Public Data plugin for ChatGPT for Healthcare. According to the company, these tools are intended to let healthcare organizations connect ChatGPT with authorized patient records and use structured information from official sources. The report does not identify specific hospitals or health systems using the integration.

The report says clinicians could use the system to review patient histories, laboratory results, medication changes and follow-up needs before appointments. It also says OpenAI described two versions of the product, ChatGPT for Clinicians and ChatGPT for Healthcare. Scripps News reports that OpenAI said 99% of responses were rated safe across more than 4,300 evaluations involving clinical use cases, but the article gives no methodology, evaluator details or underlying results.

来源详情: scrippsnews.com ↗

为什么这很重要

Connecting a general-purpose conversational AI system to clinical records could make patient context easier for authorized care teams to review, but it also raises substantial questions about privacy, accuracy, oversight and access controls. The practical value depends on how the tools handle incomplete or conflicting records and whether clinicians must verify every output. Scripps News attributes the safety results to OpenAI, and no independent evaluation is provided.

Patient-record access changes the risk profile of a chatbot. A system that summarizes authorized clinical information may reduce time spent gathering context, but an incorrect omission, outdated medication detail or misleading summary could affect care if treated as authoritative. The report does not establish that the tools make diagnoses, recommend treatment or operate without clinician review.

The announcement is significant because it places ChatGPT inside a sensitive healthcare information workflow rather than describing a general-purpose . However, the source is a secondary report of OpenAI’s claims. It does not independently confirm the integration, the safety percentage, the quality of the summaries or the privacy and security controls.

Interactive Mechanism

互动机制:它实际上是如何运作的

以交互方式探索这一发展背后的基础技术。

Agent Lifecycle Stage:
1
User Intent & Planning: "Audit customer refund request #4092 and settle payment."
2
Tool Calling: Emits structured JSON call crm_get_transaction(id='4092').
3
Guardrail & Verification:🛡️ Paused: High-value action requires human operator sign-off.
4
Final Settlement: Refund recorded, email receipt dispatched, and audit log stored.
Core takeaway: An AI agent is not just a language model—it is a closed loop of planning, tool invocation, and environment feedback. Production systems require self-healing retries and strict human approval guardrails.
交互式概念检查+10 Points
What is AI? Quiz

A route planner searches possible journeys using explicit rules. What does this illustrate about AI?

接下来看什么

The key unanswered issues are rollout timing, eligibility, pricing, geographic availability, integration requirements and the safeguards governing record access. Future reporting should establish whether the tools are available beyond selected healthcare organizations, how Epic permissions are enforced, what audit logs and human-review requirements exist, and whether independent clinical testing supports OpenAI’s reported safety results.

Scripps News does not state when the tools will launch, whether they are generally available, which organizations qualify, what regions they cover or whether access requires an enterprise agreement. It also provides no pricing information. Those conditions determine whether this is an immediately usable product or an announcement of a limited offering.

Verification should focus on permission boundaries, patient consent, data retention, auditability, handling of conflicting records, clinician sign-off and the process for correcting errors. Independent evaluations would also be needed to assess whether the reported safety rate generalizes to real clinical workflows and different patient populations.

相关指南和测验

什么是人工智能?ChatGPT 与大语言模型AI 伦理人工智能模型解释测试你所知道的——尝试免费的人工智能测验在我们的词汇表中查找人工智能术语关注 AI 模型发布跟踪器
觉得这有用吗?