Data privacy in Large Language Models: Difference between revisions

From LemonWiki共筆
Jump to navigation Jump to search
 
(12 intermediate revisions by the same user not shown)
Line 1: Line 1:
大型語言模型的資料隱私:以ChatGPT為例探討不同版本的政策
Data privacy in Large Language Models


{{Draft}}
{{Template:Generative AI Tool}}
 
{{LanguageSwitcher | content = [[Data privacy in Large Language Models | English]], [[Data privacy in Large Language Models in Mandarin | 漢字]] }}


== List of Large Language Models ==
== List of Large Language Models ==
{{Gd}} [https://claude.ai/new Claude]
 
* 除非 [https://support.anthropic.com/zh-TW/articles/7996868-%E6%88%91%E6%83%B3%E9%81%B8%E6%93%87%E9%80%80%E5%87%BA%E4%BD%BF%E7%94%A8%E6%88%91%E7%9A%84%E6%8F%90%E7%A4%BA%E5%92%8C%E7%B5%90%E6%9E%9C%E4%BE%86%E8%A8%93%E7%B7%B4%E6%A8%A1%E5%9E%8B 特殊狀況]，否則不會用作訓練<ref>[https://support.anthropic.com/zh-TW/articles/7996866-%E6%82%A8%E6%9C%83%E5%84%B2%E5%AD%98%E5%80%8B%E4%BA%BA%E8%B3%87%E6%96%99%E5%A4%9A%E9%95%B7%E6%99%82%E9%96%93 您會儲存個人資料多長時間？ | Anthropic Help Center]</ref>
[https://claude.ai/new Claude]
* Personal versions: User chat records may be used to improve AI models, but users can choose whether to participate in this training through privacy settings.<ref>[https://privacy.anthropic.com/en/articles/10023555-how-do-you-use-personal-data-in-model-training How Do You Use Personal Data in Model Training? | Anthropic Privacy Center]</ref><ref>[https://www.anthropic.com/news/updates-to-our-consumer-terms Updates to Consumer Terms and Privacy Policy \ Anthropic]</ref>
* Business versions: Conversation content is not used to train models<ref>[https://privacy.anthropic.com/en/articles/7996885-how-do-you-use-personal-data-in-model-training How do you use personal data in model training? | Anthropic Privacy Center]</ref>
 


[https://openai.com/chatgpt/ OpenAI ChatGPT]
[https://openai.com/chatgpt/ OpenAI ChatGPT]
* 個人版：個人版指不是企業版 (Enterprise)、團隊版 (Team)、API 方案，對話內容就會被拿去訓練模型 <ref>Data Controls FAQ | OpenAI Help Center https://help.openai.com/en/articles/7730893-data-controls-faq</ref><ref>"We do not train on your business data (data from ChatGPT Team, ChatGPT Enterprise, or our API Platform)" https://openai.com/enterprise-privacy/</ref>。可以前往「[https://privacy.openai.com/policies?modal=take-control OpenAI Privacy Center]」設定不要訓練模型 <ref>[https://help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance How your data is used to improve model performance | OpenAI Help Center]</ref>。
* Personal version: Personal version (non-Enterprise, non-Team, non-API plans) conversation content is used to train models <ref>Data Controls FAQ | OpenAI Help Center https://help.openai.com/en/articles/7730893-data-controls-faq</ref><ref>"We do not train on your business data (data from ChatGPT Team, ChatGPT Enterprise, or our API Platform)" https://openai.com/enterprise-privacy/</ref>. Exceptions: (1) Enabling "[https://chatgpt.com/?temporary-chat=true Temporary Chat]" will not use conversations to train OpenAI's models<ref>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."</ref>, (2) You can go to "[https://privacy.openai.com/policies?modal=take-control OpenAI Privacy Center]" to opt out of model training <ref>[https://help.openai.com/en/articles/5722486-how-your-data-is-used-to-improve-model-performance How your data is used to improve model performance | OpenAI Help Center]</ref> {{exclaim}}
* 商業版：對話內容就'''不會'''被拿去訓練模型
* Business version: Enterprise, Team, and API plan conversation content is '''not''' used to train models


[https://www.microsoft.com/zh-tw/microsoft-copilot Microsoft Copilot]
[https://www.microsoft.com/zh-tw/microsoft-copilot Microsoft Copilot]
* 免費版：對話內容就會被拿去訓練模型<ref>[https://learn.microsoft.com/en-us/copilot/privacy-and-protections Copilot Privacy and Protections | Microsoft Learn]</ref>
* Free version: Conversation content is used to train models<ref>[https://learn.microsoft.com/en-us/copilot/privacy-and-protections Copilot Privacy and Protections | Microsoft Learn]</ref>
* 企業版 (使用 Entra ID)：否
* Enterprise version (using Entra ID): No
 
[https://gemini.google.com/app?hl=en 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.<ref>[https://support.google.com/a/answer/15706919?hl=en Generative AI in Google Workspace Privacy Hub]</ref>
* '''Personal users''': Conversation content may be used to train models and reviewed by humans. '''Do not share sensitive information'''.<ref>[https://support.google.com/gemini/answer/13594961?hl=en Gemini Apps Privacy Hub - Gemini Apps Help]</ref> {{exclaim}}
* '''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<ref>[https://support.google.com/gemini/answer/13594961?hl=en Gemini Apps Privacy Hub - Gemini Apps Help]</ref>


[https://gemini.google.com/app?hl=zh-TW Google Gemini]
== Further reading ==
* 對話內容就會被拿去訓練模型 (可自行停用 Gemini 應用程式活動記錄)，人工審查員會看到你的對話內容，'''請使用者不要傳機敏資料'''。<ref>[https://support.google.com/gemini/answer/13594961?hl=en Gemini Apps Privacy Hub - Gemini Apps Help]</ref>
* [https://www.microsoft.com/zh-tw/privacy/privacystatement Microsoft 隱私權聲明 – Microsoft 隱私權]
* 資料保存時間：「相關對話內容最多會在帳戶中保留 72 小時，無論是否啟用 Gemini 系列應用程式活動記錄都一樣」<ref>[https://support.google.com/gemini/answer/13594961?hl=en Gemini Apps Privacy Hub - Gemini Apps Help]</ref>
* [https://www.microsoft.com/licensing/terms/product/PrivacyandSecurityTerms/all Microsoft Product Terms]


== References ==
== References ==
<references />
<references />


[[Category:Artificial intelligence]] [[Category:Tool]] [[Category:Generative AI]]
[[Category: Artificial intelligence]] [[Category: Tool]] [[Category: Generative AI]] [[Category: Data privacy]]

Latest revision as of 11:28, 3 September 2025

Data privacy in Large Language Models

Generative AI Tool: 📁 Prompts | 👩🏻‍💻 Text to presentation | 📢 Text to audio or speech | 🎵 Text to music | 🖼️ Text to image | 🎥 Text to video | 🚀 Coding | 🧭 Guideline | 💬 FAQ


🌐 Switch language: English, 漢字


List of Large Language Models[edit]

Claude

  • Personal versions: User chat records may be used to improve AI models, but users can choose whether to participate in this training through privacy settings.[1][2]
  • Business versions: Conversation content is not used to train models[3]


OpenAI ChatGPT

  • Personal version: Personal version (non-Enterprise, non-Team, non-API plans) conversation content is used to train models [4][5]. Exceptions: (1) Enabling "Temporary Chat" will not use conversations to train OpenAI's models[6], (2) You can go to "OpenAI Privacy Center" to opt out of model training [7] 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[8]
  • 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.[9]
  • Personal users: Conversation content may be used to train models and reviewed by humans. Do not share sensitive information.[10] 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[11]

Further reading[edit]

References[edit]

  1. ↑ How Do You Use Personal Data in Model Training? | Anthropic Privacy Center
  2. ↑ Updates to Consumer Terms and Privacy Policy \ Anthropic
  3. ↑ How do you use personal data in model training? | Anthropic Privacy Center
  4. ↑ Data Controls FAQ | OpenAI Help Center https://help.openai.com/en/articles/7730893-data-controls-faq
  5. ↑ "We do not train on your business data (data from ChatGPT Team, ChatGPT Enterprise, or our API Platform)" https://openai.com/enterprise-privacy/
  6. ↑ 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."
  7. ↑ How your data is used to improve model performance | OpenAI Help Center
  8. ↑ Copilot Privacy and Protections | Microsoft Learn
  9. ↑ Generative AI in Google Workspace Privacy Hub
  10. ↑ Gemini Apps Privacy Hub - Gemini Apps Help
  11. ↑ Gemini Apps Privacy Hub - Gemini Apps Help
Retrieved from "https://wiki.planetoid.info/index.php?title=Data_privacy_in_Large_Language_Models&oldid=25961"