Data privacy in Large Language Models

From LemonWiki共筆
Jump to navigation Jump to search

大型語言模型的資料隱私政策


icon_scale_pencil.png This article "Data privacy in Large Language Models" is still being written. If there are any incomplete parts, you are welcome to directly edit them. 這篇文章「Data privacy in Large Language Models」內容還在撰寫中,如果有不完整的部分,歡迎你直接動手修改


List of Large Language Models

Good.gif Claude

  • Personal and Business versions: Conversation content is not used to train models[1]

OpenAI ChatGPT

  • Personal version: Personal version (non-Enterprise, non-Team, non-API plans) conversation content is used to train models [2][3]. Exceptions: (1) Enabling "Temporary Chat" will not use conversations to train OpenAI's models[4], (2) You can go to "OpenAI Privacy Center" to opt out of model training [5] Icon_exclaim.gif
  • Business version: Enterprise, Team, and API plan conversation content is not used to train models

Microsoft Copilot

  • Free version: Conversation content is used to train models[6]
  • Enterprise version (using Entra ID): No

Google Gemini

  • Enterprise and Education users: Conversation content is not used to train models and is not reviewed by human reviewers, with enterprise-grade data protection.[7]
  • Personal users: Conversation content may be used to train models and reviewed by humans. Do not share sensitive information.[8] Icon_exclaim.gif
  • Data retention periods:
    • Enterprise version: Admins can control conversation retention time (auto-delete after 3, 18, or 36 months, default 18 months)
    • Personal version: When activity recording is disabled, conversation content is retained in accounts for up to 72 hours[9]

Further reading

References

  1. How do you use personal data in model training? | Anthropic Privacy Center
  2. Data Controls FAQ | OpenAI Help Center https://help.openai.com/en/articles/7730893-data-controls-faq
  3. "We do not train on your business data (data from ChatGPT Team, ChatGPT Enterprise, or our API Platform)" https://openai.com/enterprise-privacy/
  4. Website text: "This chat won't appear in your history, use or update ChatGPT's memory, or train our models. For safety reasons, we may retain a copy of this chat for up to 30 days."
  5. How your data is used to improve model performance | OpenAI Help Center
  6. Copilot Privacy and Protections | Microsoft Learn
  7. Generative AI in Google Workspace Privacy Hub
  8. Gemini Apps Privacy Hub - Gemini Apps Help
  9. Gemini Apps Privacy Hub - Gemini Apps Help