Copyright Work “15th Five-Year Plan”: China sets its 2026–2030 copyright agenda, with AI rules in the pipeline
- The National Copyright Administration issued the “15th Five-Year Plan” for Copyright Work (Guo Ban Fa [2026] No. 3) on 31 August 2026 and published it in September, setting the 2026–2030 agenda for copyright legislation, protection, services and international cooperation in China.
国家版权局于2026年8月31日印发《版权工作”十五五”规划》（国版发〔2026〕3号），2026年9月公开，确立2026—2030年中国版权立法、保护、服务与国际合作的路线图。
- The plan places AI at the centre of the next legislative cycle: it tasks the authorities with drafting rules on generative-AI copyright, building a fair-use framework for AI training data, and improving copyright rules suited to blockchain, big data and AI, while regulating copyright order in AI-related fields.
规划把人工智能置于下一立法周期核心：部署研究起草生成式人工智能版权规则、推动建立人工智能训练数据合理使用制度、完善适配区块链、大数据与人工智能发展的版权规则，并规范生成式人工智能相关领域版权秩序。
- Regulatory revisions on the horizon include the implementing regulations of the Copyright Law, the Regulations on Collective Administration of Copyright, the Regulations on the Protection of the Right of Communication through Information Networks and the Regulations on the Protection of Computer Software, plus departmental rules on copyright administrative penalties and voluntary registration.
即将推进的法规修订包括：《著作权法实施条例》《著作权集体管理条例》《信息网络传播权保护条例》《计算机软件保护条例》等行政法规，以及《著作权行政处罚实施办法》《作品自愿登记试行办法》《计算机软件著作权登记办法》等部门规章。
- Service infrastructure gets a national platform: the plan launches a project to build a nationwide unified copyright-registration information service platform, promoting blockchain-based registration pilots, online pledge (hypothecation) registration and integrated registration-query-monitoring-statistics services.
服务基础设施将建全国统一平台：规划设立”全国统一的著作权登记平台建设工程”，研究建设全国著作权登记信息服务平台，推进数字化作品著作权登记区块链运用试点、版权质权登记线上办理，实现登记、查询、监测、统计一体化服务。
- By 2030 the plan targets 9.5 million national work registrations, 4 million computer-software copyright registrations and RMB 12 billion in registered copyright pledge financing; enforcement actions (including the “Sword Net” anti-online-piracy campaign) will keep targeting books, films, software, live streaming and streaming media, with copyright order in generative-AI use added to the list.
规划提出到2030年全国作品登记数量950万件、计算机软件著作权登记400万件、版权质权融资登记金额120亿元；执法行动（含”剑网”打击网络侵权盗版专项行动）将持续覆盖图书、影视、软件、网络直播与流媒体等领域，并把生成式人工智能相关领域版权秩序纳入治理范围。
- For foreign right holders, the five-year signals are practical: digital registration infrastructure will make Chinese copyright records easier to obtain and verify; software and AI-trainers operating in China face clearer licensing expectations for content corpora; and the collective-management and trading-platform agenda will shape how foreign repertoires are licensed and paid in China.
对外国权利人的五年信号很具体：数字化登记基础设施将使中国版权记录更易取得与验证；在华运营的软件与人工智能训练方在语料许可上将面对更清晰的预期；集体管理与交易平台议程将塑造外国曲目与内容在中国如何获授权与获得报酬。
Copyright Work “15th Five-Year Plan”: China sets its 2026–2030 copyright agenda, with AI rules in the pipeline | 《版权工作”十五五”规划》印发：中国2026—2030版权路线图与AI规则前瞻
Overview
In late August 2026 the National Copyright Administration (NCA) issued the “15th Five-Year Plan” for Copyright Work (《版权工作”十五五”规划》, Guo Ban Fa [2026] No. 3) and published it in early September. The plan sets the direction of China’s copyright system for 2026–2030 across five task areas: completing the socialist copyright legal system with Chinese characteristics; building an innovation-friendly copyright-protection ecosystem; establishing efficient and standardised copyright service mechanisms; shaping an open and win-win pattern of international copyright cooperation; and fostering a society that respects copyright.
The document matters well beyond China’s borders. It is the clearest official statement to date of how China intends to handle copyright in the era of generative AI, and it maps the infrastructure — registration, collective management, trading platforms and pledge finance — through which foreign works, software and repertoires are licensed and monetised in the Chinese market.
The AI agenda moves to the centre
The most significant content for technology companies is the treatment of AI across the plan:
- Generative-AI copyright rules are to be drafted. The plan’s regulatory-revision project explicitly lists “research and drafting of generative-AI copyright rules”.
- A fair-use framework for AI training data is to be built. The plan calls for “promoting the establishment of a fair-use system for AI training data” (推动建立人工智能训练数据合理使用制度), a first for an official planning document at this level.
- Copyright rules are to adapt to blockchain, big data and AI, with protection systems refined for online literature, games, audiovisual content, and basic and industrial software.
- Copyright order in generative-AI fields is to be regulated, and AI-related copyright compliance problems actively addressed — signalling enforcement attention to unauthorised use of works in model training and output.
- Collective-management organisations are assigned a role in licensing works for AI training-data corpora and in managing the risk of abusive mass litigation.
The legislative and regulatory pipeline
The plan’s “copyright laws and regulations revision project” lists the next wave of rule-making:
- Administrative regulations: the implementing regulations of the Copyright Law, the Regulations on Collective Administration of Copyright, the Regulations on the Protection of the Right of Communication through Information Networks, and the Regulations on the Protection of Computer Software.
- Departmental rules: the Measures on Administrative Penalties for Copyright Infringement, the Interim Measures on Voluntary Registration of Works, and the Measures on Computer-Software Copyright Registration.
- Treaty implementation: effective implementation of the Beijing Treaty on Audiovisual Performances and the Marrakesh Treaty.
Foreign right holders holding Chinese software, film, music, book or database assets should read this pipeline as a commitment to keep updating the implementing layer of the 2021 Copyright Law, including the provisions most relevant to digital exploitation.
Registration and service infrastructure
A distinct strand of the plan is service modernisation:
- A project to build a nationwide unified copyright-registration information service platform, using blockchain, big data and AI to raise the level of digital registration, and realising integrated registration, query, monitoring and statistics services.
- Pilots for blockchain-based registration of digital works, encouraging registration of data that constitutes a work.
- An online platform for copyright pledge registration and an information platform for publicising copyright pledge registrations, moving pledge finance online.
- Continued development of a unified online platform for copyright administrative enforcement.
The headline 2030 targets are concrete: 9.5 million registered works nationwide, 4 million computer-software copyright registrations, and RMB 12 billion in registered copyright-pledge financing. For foreign companies that register works in China — whether for legal protection or for platform takedown processes — the platform upgrade should shorten timelines and make certificates easier to verify.
Enforcement and the protection ecosystem
The plan’s protection agenda retains the familiar campaign structure: the “Sword Net” campaign against online piracy, the “Sword Shadow” campaign for cinema-film copyright protection, and the youth copyright-protection season. Priority industries include book publishing, radio, television and online audiovisual, online literature, online games and online music; priority products include textbooks and teaching materials, bestsellers, cinema films, major sports events and cultural-creative products; and priority channels include online sales, online storage and streaming-media distribution. Generative-AI-related copyright order is added to the governance list.
What this means for foreign right holders
- Registration becomes easier and more meaningful. With a unified national platform, blockchain pilots and clearer rules, foreign right holders registering works in China should expect faster processing and stronger evidentiary value for takedowns, litigation and customs protection.
- AI and software companies face clearer licensing expectations. If your operations in China train models on Chinese-language corpora or distribute software that reproduces third-party content, expect the fair-use framework to define what needs a licence, and structure data-acquisition contracts accordingly.
- Collective management will matter more. With collective-management organisations given an explicit role in licensing content for AI training and in curbing abusive litigation, users of music, text and audiovisual repertoires should confirm which society covers the works they use.
- Pledge finance becomes usable. Online pledge registration and the RMB 12 billion target signal a functioning copyright-financing channel — relevant for IP-rich subsidiaries of foreign groups seeking asset-backed financing in China.
- Treaty-based protection. Continued implementation of the Beijing and Marrakesh Treaties underpins protection for audiovisual performers’ rights and accessible-format copies, relevant to distributors of performances and of accessible content.
Practical steps
- Map your China work registrations (literary, software, artistic, audiovisual) and plan to migrate records onto the unified platform as modules come online during the 15th Five-Year Plan period.
- If you license or use content corpora for AI training or software features, review contracts against the coming fair-use framework and add licence-reservation clauses.
- Confirm the collective-management coverage of the music and text repertoires you use in China, and budget for licence fees under the reformed collective-administration rules.
- For software distributed in China, keep copyright registration current — registration remains the practical gateway for administrative complaints and platform enforcement.
- Track the drafting of the generative-AI copyright rules; when published for comment, consider submitting industry input on training-data licensing and output liability.
Sources
- 国家版权局官网 — 国家版权局关于印发《版权工作”十五五”规划》的通知（国版发〔2026〕3号）: https://www.ncac.gov.cn/xxfb/tzgg/202609/t20260907_1007129.html
- 中国政府网政策库 — 版权工作”十五五”规划全文: https://www.gov.cn/zhengce/zhengceku/202609/content_7080386.htm
- 中国共产党新闻网（人民日报2026年9月8日）— 《版权工作”十五五”规划》印发: https://cpc.people.com.cn/BIG5/n1/2026/0908/c64387-40794421.html
Related reading
- see also: China’s overall 15th five-year plan for IP protection and utilisation (ip-15th-five-year-plan-2026)
- see also: how foreign firms handle software copyright and registration in China (software-copyright-foreign-firms)
《版权工作”十五五”规划》印发：中国2026—2030版权路线图与AI规则前瞻
概述
2026年8月底，国家版权局印发《版权工作”十五五”规划》（国版发〔2026〕3号），并于2026年9月初对外公开。规划围绕五个任务领域确立2026—2030年中国版权制度方向：完善中国特色社会主义版权法律制度；建设激励创新的版权保护生态；建设规范高效的版权服务机制；建设开放共赢的版权国际合作新格局；建设尊重版权的社会环境。
这份文件的份量远超国境。它迄今最清晰地表明中国拟如何在生成式人工智能时代处理版权问题，并描绘了外国作品、软件与曲目在中国市场获得授权与变现所依托的基础设施——登记、集体管理、交易平台与质权融资。
AI议程走向中心
对科技企业而言，最具含金量的是规划中人工智能的处置方式：
- 研究起草生成式人工智能版权规则。 规划的法规修订工程明确单列”研究起草生成式人工智能版权规则”。
- 推动建立人工智能训练数据合理使用制度。 规划提出”推动建立人工智能训练数据合理使用制度”——在此层级的官方规划文件中尚属首次。
- 完善适配区块链、大数据与人工智能的版权规则，并完善网络文学、网络游戏、网络视听以及基础软件、工业软件版权保护制度。
- 规范生成式人工智能相关领域版权秩序，积极推动解决人工智能相关版权合规问题——释放对未经授权将作品用于模型训练与输出的执法关注信号。
- 集体管理组织被赋予角色：在人工智能训练数据语料库作品许可中发挥作用，并治理批量维权诉讼风险。
立法修法管线
规划的”版权法规规章修订工程”列出下一波制度建设：
- 行政法规：《著作权法实施条例》《著作权集体管理条例》《信息网络传播权保护条例》《计算机软件保护条例》。
- 部门规章：《著作权行政处罚实施办法》《作品自愿登记试行办法》《计算机软件著作权登记办法》。
- 条约实施：推动《视听表演北京条约》《马拉喀什条约》的有效实施。
在华持有软件、影视、音乐、图书或数据库资产的外国权利人，应把这条管线理解为持续更新2021年著作权法实施层、尤其是数字利用相关条款的承诺。
登记与服务基础设施
规划的另一条主线是服务现代化：
- 设立”全国统一的著作权登记平台建设工程“，研究建设全国著作权登记信息服务平台，充分利用区块链、大数据、人工智能等技术提升登记数字化水平，实现登记、查询、监测、统计一体化服务。
- 推进数字化作品著作权登记区块链运用试点，鼓励对构成作品的数据开展著作权登记。
- 建设版权质权登记公示信息平台，推进质权登记线上办理。
- 完善版权行政执法一体化平台。
2030年核心指标具体明确：全国作品登记数量950万件、计算机软件著作权登记400万件、版权质权融资登记金额120亿元。对在华办理作品登记的外国企业——无论为法律保护还是平台下架流程——平台升级应缩短周期并使证书更易验证。
执法与保护生态
规划的保护议程延续熟悉的行动框架：”剑网”打击网络侵权盗版专项行动、”剑影”院线电影版权保护专项行动、青少年版权保护季。重点产业包括图书出版、广播电视和网络视听、网络文学、网络游戏、网络音乐；重点产品包括教材教辅、畅销图书、院线电影、重大体育赛事、文创产品；重点渠道包括网络销售、网络存储、流媒体传播。生成式人工智能相关领域版权秩序被纳入治理范围。
对外国权利人的意义
- 登记更便利、更有意义。 依托全国统一平台、区块链试点与更清晰的规则，在华登记作品的外国权利人应可期待更快处理、以及更强的下架、诉讼与海关保护证据价值。
- AI与软件公司面对更清晰的许可预期。 若在华运营涉及以中文语料训练模型，或分发复制第三方内容的软件，公平使用框架将界定何种情形需要许可——应据此安排数据采购合同。
- 集体管理将更重要。 集体管理组织被明确赋予”为AI训练许可内容、遏制滥诉”的角色，使用音乐、文字与视听曲库的主体应确认所用作品的覆盖管理组织。
- 质权融资可用化。 质权登记线上化与120亿元目标，意味着可运行的版权融资渠道——对寻求以知识产权资产融资的外国集团中国子公司具有意义。
- 条约保护。 北京条约与马拉喀什条约的持续实施支撑表演者权与无障碍格式复制件的保护，与演出内容及无障碍内容的分发者相关。
实务建议
- 梳理中国作品登记资产（文字、软件、美术、视听），并计划在”十五五”期间各功能模块上线时将记录迁移至统一平台。
- 若为AI训练或软件功能而许可或使用内容语料，对照即将形成的公平使用框架复核合同，并增加授权保留条款。
- 确认在华使用的音乐与文字曲库的集体管理覆盖情况，按改革后的集体管理规则为许可费做预算。
- 对在华分发的软件，保持著作权登记更新——登记仍是行政投诉与平台执法的实务入口。
- 跟踪生成式人工智能版权规则的起草；发布征求意见稿后，可考虑就训练数据许可与输出责任提交行业意见。
来源
- 国家版权局官网 — 国家版权局关于印发《版权工作”十五五”规划》的通知（国版发〔2026〕3号）: https://www.ncac.gov.cn/xxfb/tzgg/202609/t20260907_1007129.html
- 中国政府网政策库 — 版权工作”十五五”规划全文: https://www.gov.cn/zhengce/zhengceku/202609/content_7080386.htm
- 中国共产党新闻网（人民日报2026年9月8日）— 《版权工作”十五五”规划》印发: https://cpc.people.com.cn/BIG5/n1/2026/0908/c64387-40794421.html
相关阅读
- 见：知识产权保护和运用”十五五”规划（ip-15th-five-year-plan-2026）
- 见：外国企业的软件著作权与中国登记实务（software-copyright-foreign-firms）