📄 中文摘要
在使用 GitHub Copilot 进行实际项目开发时,尤其是在公司环境中,隐私问题显得尤为重要。开发者在编辑器中输入的是专有逻辑,涉及内部 API、客户工作流程和关键业务算法,同时又连接到一个基于云的 AI 系统,该系统根据发送的代码上下文生成建议。这引发了一系列关于代码存储、是否用于训练模型、其他用户是否能看到类似私有逻辑以及用户实际控制权的问题。GitHub Copilot 根据用户的使用方式和所选择的计划提供不同的隐私保障和治理措施。
📄 English Summary
Privacy options available in GitHub Copilot
Using GitHub Copilot in real projects, especially within a corporate environment, raises significant privacy concerns. Developers input proprietary logic into their editors, working with internal APIs, customer workflows, and critical business algorithms while connected to a cloud-based AI system that generates suggestions based on the code context provided. This situation leads to serious questions regarding code storage, usage for model training, visibility of private logic to other users, and the actual controls available to users. The privacy guarantees and governance measures offered by GitHub Copilot vary depending on how it is used and the plan selected.
Powered by Cloudflare Workers + Payload CMS + Claude 3.5
数据源: OpenAI, Google AI, DeepMind, AWS ML Blog, HuggingFace 等