Recently, the rapid development of generative AI has once again brought copyright, data compliance, and content governance to the forefront of global industry discussions. Companies like OpenAI and Meta frequently face copyright lawsuits from creator associations and publishing institutions, while regulatory bodies in the United States, Europe, and China are intensifying efforts to implement policies regarding the legality of training data and copyright protection. These events reveal a core contradiction: AI models rely on vast amounts of content for training, while traditional copyright systems struggle to keep pace with the speed and scale of technological advancements.
For example, in June 2024, the Recording Industry Association of America sued two AI music companies—Suno and Udio—accusing them of infringing music copyrights while training their AI tools. Additionally, in December 2023, The New York Times announced it had filed a lawsuit in the Southern District of New York, accusing Microsoft and OpenAI of copyright infringement and misuse of the media's intellectual property.
In China, on June 20, 2023, the Beijing Internet Court heard the country's first case involving copyright infringement related to the training of AI painting models. Four illustrators sued the company behind the AI painting product "Trik AI," owned by Xiaohongshu, and its platform company, Xingyin Technology, accusing them of using the plaintiffs' artistic works without authorization to train AI models and generate highly similar images, constituting copyright infringement.
These cases highlight the contradiction between AI and traditional copyright systems: AI models depend on vast amounts of content for training, while traditional copyright systems struggle to keep pace with the speed and scale of technological advancements. As noted by the Harvard Business Review, the proliferation of generative AI has brought unprecedented intellectual property issues, eroding creators' rights while also exposing companies to legal uncertainties.
In this context, recent research by Messari has proposed the concept of an IP Layer, emphasizing the digitization of intellectual property into structured, verifiable on-chain assets to provide compliant training and invocation data for AI. The application scenarios depicted in the report show that once music, video, or image works are registered on-chain, they can serve not only as proof of copyright but also be legally invoked by AI models, with revenue automatically settled through smart contracts. This approach effectively addresses the issue of "gray data" in generative AI training while providing a transparent and traceable revenue mechanism for the creator economy. For instance, if a photographer registers their photos on-chain, AI models can legally invoke these photos to generate new visual works, while automatically paying the photographer usage fees through smart contracts, without manual intervention.
Technically, the implementation of the IP Layer involves several key components. The Origin framework is responsible for content on-chain registration, rights confirmation, and transactions, combined with a dispute arbitration mechanism to provide creators with a traceable and verifiable digital identity. Meanwhile, mAItrix provides the execution environment for the AI invocation layer, including Retrieval-Augmented Generation (RAG), Trusted Execution Environment (TEE), and smart profit-sharing logic, allowing AI to safely and efficiently access IP data under compliant conditions. This architecture not only reduces legal risks for AI developers but also addresses performance and cost issues related to the invocation of vast amounts of content on-chain. RAG technology allows AI to generate novels or advertising copy by only invoking authorized text segments on-chain, rather than all data on the internet, ensuring legality and compliance; TEE acts like a "cryptographic black box," ensuring that data invoked by AI cannot be stolen or tampered with externally.
Technical Depth Expansion: Smart Contracts, On-Chain Storage, and AI Invocation Optimization
In the implementation of the IP Layer, the design of smart contracts is the core driving mechanism. The profit-sharing logic automatically tracks each IP invocation or transaction through on-chain contracts and distributes revenue to creators, platforms, and AI developers according to pre-set ratios. For example, when an AI model uses a segment of music or video for generation, the smart contract records the number of invocations, timestamps, and invoking party information, triggering revenue settlement in real-time. This not only achieves transparency and traceability but also reduces manual operation costs and legal risks. If a short video AI model invokes a certain on-chain music segment 10 times, the smart contract automatically calculates the profit-sharing and distributes the revenue to the music creator and video platform without manual processing.
In terms of on-chain data storage, to cope with vast amounts of content and high-frequency invocations, the IP Layer typically employs a layered storage and indexing mechanism. Core IP metadata (creator identity, copyright information, authorization conditions) is stored on-chain to ensure immutability and verifiability; large file content (such as videos and audios) is stored in decentralized storage systems (like IPFS or Filecoin), combined with encrypted indexing for rapid retrieval. Through this architecture, the system ensures both security and compliance while meeting the performance demands of high-frequency AI invocations. The large file content of a movie is stored on IPFS, but its copyright information, authorization conditions, and revenue rules are recorded on the blockchain. When AI generates recommended clips or promotional materials, it only needs to query on-chain information to legally invoke movie segments.
The optimization of the AI invocation layer is equally crucial. The RAG (Retrieval-Augmented Generation) mechanism allows AI to retrieve on-demand data from on-chain or decentralized storage when generating content, reducing data redundancy before training and improving generation efficiency. Combined with the Trusted Execution Environment (TEE), the data during the AI invocation process is encrypted throughout the computation, ensuring content security and privacy. Smart contracts record invocation and profit-sharing information in real-time, allowing for immediate revenue settlement for creators while forming an auditable invocation log that provides a chain of evidence for potential legal disputes. In educational scenarios, AI can invoke authorized course videos on demand to generate customized learning materials, with the revenue generated from student payments automatically distributed to course creators and platforms.
In practical terms, the IP Layer has already shown potential in certain vertical industries. Reports from Messari and WIPO indicate that scenarios such as music remixing, short videos, and educational content are among the first to become testing grounds. Through on-chain registration and smart contract settlement, creators can receive immediate revenue, while AI can invoke this data for training or generating new content, thus forming a closed loop of content on-chain, invocation, and profit-sharing.
At the same time, the IP Layer also presents new ecological and governance challenges. The copyright disputes of generative AI are characterized by subjectivity and cross-jurisdictional features, making it difficult to resolve solely through smart contracts; therefore, it requires a combination of community governance, professional arbitration, and on-chain automation logic. The "opt-in" model and best practices for copyright security proposed by EU commissioned research also provide policy references for the implementation of the IP Layer: AI invocation must obtain explicit authorization rather than being assumed available by default, thereby institutionally safeguarding creators' rights.
Future Evolution and Ecological Expansion: Cross-Chain, Interoperability, and New Business Models
Looking ahead, the potential of the IP Layer lies not only in copyright protection but also in its capacity to expand as a global content infrastructure. The application of cross-chain technology allows IP data from different blockchain ecosystems to interoperate, enabling works registered by creators on one platform to be invoked by AI on another platform while maintaining copyright and revenue tracking. AI models can also access on-chain data across platforms, achieving global integration of training data, thereby improving model accuracy and diversity.
The global content platform integration is another key dimension of the IP Layer's expansion. Music, film, education, and research platforms can connect to the IP Layer through a unified on-chain API and smart contracts, enabling cross-platform authorization, revenue settlement, and data invocation. This interoperability not only reduces platform integration costs but also enhances the transparency and sustainability of creator revenues.
Moreover, the IP Layer also fosters new business models. For instance, AI can rent on-chain data for training on demand, forming a fee model similar to cloud computing resources; content creators can earn revenue simultaneously across multiple platforms, creating a "cross-platform copyright circulation network"; decentralized arbitration and community governance mechanisms may also give rise to new services and value-added products, such as on-chain copyright insurance or content legality certification services. These models not only transform the traditional content industry chain but also provide new value opportunities for investors and developers.
In practical cases, Camp Network has become a typical representative. Its Origin + mAItrix dual-layer architecture provides a complete link from IP registration to AI invocation, with millions of IP assets registered during the testnet phase and collaborations with content platforms underway. As the token generation event (TGE) approaches, the market is focused on whether it can implement the IP Layer concept, promote cross-platform ecological development, and create a sustainable value network for creators and AI developers.
Overall, the IP Layer represents a new type of infrastructure at the intersection of AI and blockchain. It not only addresses the urgent needs for copyright protection and data compliance but also provides actionable solutions for the creator economy, AI data supply chain, and decentralized governance. As technology matures, more implementation cases emerge, and cross-chain and cross-platform ecosystems expand, it is expected to reshape the value circulation model of the content industry, making intellectual property a foundational infrastructure in the AI era rather than a static rights declaration.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
