OpenAI 行动
287 人已加入OpenAI 行动全部成员
已验证邮箱并加入该行动的账户数；不表示法律资格已核实。
面向认为自己的作品可能被 OpenAI 产品（例如 ChatGPT）使用的人。找到同类人，了解已知情况，知道自己可以先做什么。
接下来怎么做
选择你的作品类型，看对应的指南。这些分类只用于筛选内容，不会产生分组人数。
指南：你可以自己做什么
- 材料清单：E 编号是什么意思
- 记录底稿：11 份可以照抄的空白表
- SOP01｜保留原作品：文字、书、绘画、照片
- SOP02｜先保住眼前的现场
- SOP03｜保存完整输入、输出和相关会话
- SOP04｜制作原作与输出对照
- SOP05｜生成指纹、记保存过程并安全备份
- SOP06｜整理既有发表、授权和权利记录
- 插画与美术：你可以先保存什么
另有其他语言版本 (9)
- The material list: what the E numbers mean (英语)
- Record templates: ten blank notes you can copy (英语)
- SOP01 | Preserving the original work: writing, books, illustration, photographs (英语)
- SOP02 | Preserve what is in front of you first (英语)
- SOP03 | Save the full input, output and related conversation (英语)
- SOP04 | Build a comparison between the work and the output (英语)
- SOP05 | Produce fingerprints, record the process and back it up safely (英语)
- SOP06 | Organise existing publication, licence and rights records (英语)
- Illustration and visual art: what you can keep (英语)
各地的规则
和你有同样疑问的人不止一个。这些页面写的是你所在地的规则实际怎么说，每条都附来源。是背景资料，不是法律建议。
- 加拿大：规则概览
- 中国内地：规则概览
- 西班牙：规则概览
- 欧盟：规则概览
- 印度：规则概览
- 墨西哥：规则概览
- 美国：规则概览
- Canada: what the rules say (英语)
- China (mainland): what the rules say (英语)
- Spain: what the rules say (英语)
- European Union: what the rules say (英语)
- India: what the rules say (英语)
- Mexico: what the rules say (英语)
- United States: what the rules say (英语)
案件与进展
- AI Copyright Deal: Albanese and Taylor Move Toward a Licence-Free Model
- Australian panel urges PM to tighten AI safeguards after OpenAI breaches; calls for creator-focused copyright review
- UK privacy watchdog secures AI company commitments on how personal data is used to train AI models
- Penguin Random House verklagt OpenAI wegen ChatGPT-„Memorisierung“ von „Der kleine Drache Kokosnuss“
- Carlsen Verlag, Marc-Uwe Kling und Astrid Henn verklagen OpenAI Ireland wegen „Das NEINhorn“ vor dem LG München I
- Leaked Australian proposal would allow AI training opt-out, with no compensation under one option
- OpenAI publishes crawler controls: GPTBot training opt-out via robots.txt
- U.S. Justice Department files statement of interest backing fair use for AI training in New York Times case
相关参考（不涉及 OpenAI）
- Australia rules out AI 'text and data mining' copyright exception, sides with creators — Frontier-model trainer whose AI-training scraping practices are the target of Australia's rejected TDM exception; decision affects it as a developer of models trained on copyrighted works. No company named as defendant; linked as reference to the policy decision.
- AI Lawsuit Tracker counts 248 cases against 76 companies; copyright dominates — Named as the most-sued company in the tracker (55 cases).
- Anthropic says new safeguards cut AI model-distillation attacks by 90–99% — Named as one of the leading labs that share Anthropic's concern about distillation letting rivals copy model capabilities.
- EFF guide: how to opt out of AI training on ChatGPT, Gemini, Claude, Copilot and other major platforms — Guide links to ChatGPT's own documentation for opting out of training on conversations.
- European Commission opens consultation on copyright and AI, asking whether rules on protected content in AI need strengthening — Major general-purpose AI provider placing models on the EU market; the consultation seeks input from generative AI providers and reviews the copyright framework governing their use of protected content. No single company is the subject.
讨论
浏览讨论不需要登录；发帖需要登录并加入本行动。
已核验律师发布
还没有已核验律师在这里发布内容。