The wave of generative AI is reshaping the content industry. Over the past two years, tools like ChatGPT, Midjourney, Suno, and Runway have made it clear that the speed and cost of creation are being exponentially reduced. With just a single prompt, songs, illustrations, images, and even short videos can be generated. AI has become not only an "assistant" but, in some scenarios, even a direct "creator." Behind this seemingly boundless creative landscape lies an increasingly pressing contradiction: Where does the training data powering these models come from? Is it legally licensed? Can the creators of the data being used receive reasonable compensation? This is no longer simply a legal or ethical issue; it's a crucial question directly impacting the future structure of the content industry.

Looking back at the copyright system over the past few decades, we find that its operating logic has fundamentally relied on registration, contracts, and legal proceedings. This may have been feasible in the era of paper publications, physical records, and film and television works, but today, faced with the massive amounts of data required for AI training, this mechanism has almost completely failed. Thousands of songs, hundreds of millions of images, and countless text fragments are silently absorbed into the model's training library, and the authors behind them can neither prevent nor benefit from this. This mismatch not only erodes the enthusiasm of creators but also puts AI products themselves in a potential compliance crisis. Several ongoing lawsuits in the United States—such as the lawsuit filed by writers against OpenAI and Getty Images' lawsuit against Stability AI—have demonstrated that copyright disputes are no longer fringe issues but core risks that will directly affect the commercialization path of AI companies.

It was against this backdrop that I came across Messari's report, "Camp Network: The New IP Layer." The report proposes a highly insightful concept: the IP Layer, or "a new foundational layer for intellectual property." Its core vision is to make IP a natively digital asset, capable of on-chain registration, tracking, and transfer, naturally endowed with ownership confirmation and transfer capabilities, much like Bitcoin or Ethereum. This vision made me realize that we're nearing a restructuring of institutional structures. While intellectual property rights used to rely on external systems and institutions for verification, in the future they may be embedded directly into code, automating ownership confirmation, authorization, and profit sharing.

Camp Network is an early experiment in this approach. It proposes starting with gas-free registration, lowering the barrier to entry for creators. Then, through the Origin framework, it enables on-chain copyright confirmation and circulation of works. Finally, it collaborates with music and short video platforms to implement specific application scenarios. Music and short video are not random tracks, as they are the areas most prone to copyright conflicts. Countless musicians have complained about their works being reused on TikTok without compensation, and countless short video creators have found their content being scraped and trained by AI models without even being credited. Camp attempts to answer these questions with on-chain copyright confirmation and contractual logic, automating content registration, usage, and revenue distribution. While it may not completely disrupt the existing system in the short term, it has demonstrated that IP ownership and circulation can be completely redesigned, without relying solely on legacy laws and platforms.

Beyond Camp, there are other equally important explorations. Story Protocol is positioned more like an IP operating system, encouraging creators to not only register their works but also allow others to create derivative works on the blockchain. The revenue from all derivative works can be tracked and automatically distributed. This will allow the future content ecosystem to increasingly resemble the logic of open source software, with each new work potentially becoming the foundation for the next. This not only lowers the barrier to collaboration but also fosters a network effect within the creation itself. Numbers Protocol takes a different approach, focusing not on copyright ownership but on the authenticity of content. In an era of AI-generated imagery and deepfakes, determining the provenance of a piece of news footage or a photograph has become a challenge that society must confront. Numbers uses blockchain to establish a traceable record for each piece of content, ensuring immutability from the device it was created on to the transmission path. This has made it particularly attractive to the media and art industries, as authenticity itself represents a new form of scarcity. KOR Protocol is more niche, targeting the music industry. The music industry has long been constrained by a complex network of copyright agencies, resulting in inefficient transactions and opaque revenue distribution. Through on-chain registration and contract authorization, KOR allows DJs and producers to directly use other people's music and instantly settle profits, bypassing traditional intermediaries. Bittensor, on the other hand, is more like the other side of the AI incentive network. Rather than addressing copyright issues, it builds a decentralized network where developers contribute models and computing power to training and receive incentives through tokens. It complements Camp's logic: the former focuses on the legality of data, while the latter focuses on who performs training and how profits are distributed. If the two can be integrated in the future, they may form a complete decentralized AI ecosystem, from data to models to applications.

These projects present different perspectives, but they all respond to the same proposition: in the AI era, intellectual property needs to be redefined. It is no longer just a concept attached to a contract, but needs to have digital native characteristics that can be registered, combined, tracked, and automatically distributed. This is not only an efficiency issue, but also an issue of industrial structure. Imagine if every song, every text, and every frame of image could exist as an on-chain asset in the future, then the relationship between creators and users would be reshaped. AI companies will no longer be "free to grab" but must purchase legal and traceable data just like purchasing computing power. This will give rise to a huge data market, making training data an infrastructure like electricity or bandwidth.

However, this road is not smooth. First, there's the issue of legal recognition. Whether on-chain content ownership can be recognized as valid in the real-world legal system remains to be seen. Second, there's the issue of user experience. Most creators are unfamiliar with on-chain operations, and if the barrier to entry is too high, they might not be willing to use them. Furthermore, the attitude of major platforms is crucial. If platforms like Spotify, YouTube, and TikTok refuse access, it will be difficult for an on-chain content ownership ecosystem to scale. Finally, even if on-chain content ownership is achieved, we must be wary of the emergence of new "on-chain oligopolies." Whether creators can truly receive reasonable returns, or whether they will be squeezed out again by new technological intermediaries, remains an unresolved issue.

Notably, the capital market has keenly grasped this trend. Story Protocol secured $36 million in funding in 2023, with investors including a16z crypto, demonstrating that top venture capital firms view the IP layer as a key development path for the next decade. Meanwhile, major AI companies are accelerating their search for compliance pathways. OpenAI has reached licensing agreements with news organizations, Google has entered into negotiations with music copyright holders, and Adobe has launched Firefly, a generative AI based on licensed content. All of these actions demonstrate that even the most powerful AI giants are forced to confront the issue of "data provenance." Under the dual pressures of law and market forces, these companies may become key partners for on-chain IP platforms.

From a longer-term perspective, the assetization of digital IP could propel the entire content industry toward a logic similar to that of the financial market. Works are no longer simply consumer goods; they become assets that can be split, transferred, and pledged. Creators can earn ongoing income through licensing, while investors can participate in profit sharing by purchasing IP assets. Platforms play a role in matchmaking and oversight. Once this ecosystem takes shape, the boundaries between the content and financial industries will be significantly broken down. Just as the internet reshaped the logic of information dissemination, the combination of AI and blockchain is reshaping the logic of creation and copyright confirmation.

Therefore, from Camp to Story, Numbers, KOR, and Bittensor, we're not witnessing a sporadic succession of projects, but rather a brewing structural shift. Each emphasizes a different dimension, yet they may one day form a cohesive whole with a clear division of labor. Camp focuses on infrastructure, Story emphasizes composability, Numbers protects authenticity, KOR specializes in the music industry, and Bittensor focuses on AI incentives. They all point to a core truth: in the AI era, whoever owns the data, controls the rights, and designs the distribution rules will control the future landscape of the content industry. Camp may not be the only answer, but its exploration shows us that IP rights can be completely reinvented. Perhaps this is the biggest variable in the content industry over the next decade.