怎么保护自己的数据How to Protect Your Data
微型指南 · 约 6 分钟Guide · about 6 min · 引用核验 2026-11-03（仅引用链接与日期） · refs checked 2026-11-03 (reference links and dates only)
把资料交给 AI 之前，先想清楚它会去哪里、会留下什么。这份指南讲六个动作。Before handing material to an AI, be clear about where it goes and what it leaves behind. Six moves.
先分清资料能不能出门：个人身份、公司机密、客户数据各算一类。
看不懂的条款别默认同意，重点看数据会不会用于训练。
能匿名化就先匿名化：去掉姓名、编号、联系方式再贴进去。
敏感场景可评估本地或私有部署；仍要检查模型接口、遥测、插件和日志是否向外发送数据。
用完检查留存：对话记录和上传文件能不能删、多久删。
给团队定规矩：哪些资料能进 AI，哪些必须走内部工具。
Sort material by whether it may leave: personal identity, company secrets, and customer data are three different categories.
Do not accept terms you cannot read; check specifically whether your data is used for training.
Anonymize first: strip names, IDs, and contact details before pasting anything in.
Assess local/private deployment for sensitive work, while checking outbound model calls, telemetry, plugins and logs.
Afterwards, check retention: can the chat history and uploaded files be deleted, and how soon?
Set team rules: which materials may go into an AI, and which must stay in internal tools.
例子与判断Example and interpretation
以下是教学情境，不是本站运行或实测记录。A teaching scenario, not a measured or executed site test.
准备让工具整理客户反馈时，先复制一份练习材料，把姓名换成客户 A、订单号换成示例编号。再检查自由文本里是否仍有地址或联系方式。替换几个字段不代表整份材料已经匿名化。Before summarizing customer feedback, make a practice copy and replace names and order numbers. Then inspect free text for addresses or contact details. Replacing a few fields does not establish full anonymization.
隐私与合规Privacy & Compliance 数据不出网Data Sovereignty 数据驻留Data Residency