This is an experimental essay written in collaboration with ChatGPT. It is intended to be exploratory, future-looking, and thought provoking and not to promote adoption or suggest legal advice.
This year as a Berkman Klein Center affiliate, I’ve been spending my time thinking about the future of the book and researching how technology might help foster publishing ecosystems with many winners. With important law suits in swing concerning the future of digital lending, now is an important time for such research to develop.
As the digital book market continues to grow, a tension may be noted between the needs of publishers for control and security and of customers for accessibility and equity. While leasing models are preferred by publishers to control the reading experience and prevent mass duplication, this approach is less favorable to many customers who wish to retain the same ownership and usability rights they enjoy with the physical books they purchase. While historical techniques like DRM may deter mass copying, it may do so by unproductively limiting the accessibility of e-books for those who depend on certain reading affordances, like screen readers. This proposal explores a model that may placate publishers with additional safeguards against mass sharing, while enabling ownership and minimizing unnecessary limitations for customers, all while furthering the public domain and the spirit of copyright.
A major consideration for such a solution is that it shouldn’t require publishers to implement or adopt costly changes to their existing systems. Instead, such a system would preferably be one a publisher could opt-in to in order to make publication easier, cheaper, and reach broader audiences. For customers, ownership of their products shouldn’t be contingent on the ongoing survival of individual vendors or publishers but rather safely reside in a reliable, long-term storage repository run by a centralized, trusted organization like the Library of Congress or a decentralized blockchain. Such a system, if it is to support one’s first sale doctrine rights of ownership, should enable customers to access their content on any number of approved, compatible applications, given these applications implement the appropriate DRM standards and that access to the files are limited solely to the designated recipient(s) with access keys. Any number of publishers, vendors, and customers would be able to mutually participate in such a system without concerns of power asymmetry or rule changes after sale.
Here’s a rough sketch of how such a system might work. Each customer has a RSA 2048 key pair consisting of a public and private key. This key pair should be consequentially tied to their identity as a wallet*; something a customer has a vested interest in protecting, akin to reputation, a credit score, or SSN. This is both what may give sellers confidence in selling to a customer and also what may deter a customer from sharing their keys (or the products to which they they grant access) to another. Because customers’ identities are consequentially tied to their secure private keys, this discourages unauthorized sharing because customers who give away or share their keys may also be giving away access to their wallets or compromising their sensitive information.
*Author’s note: Significant regulatory work should be done to ensure such a reputation system isn’t used in discriminatory ways. The primarily purpose intended in this design is to track abuse or complaints.
Publishers produce a work which they encrypt using an agreed upon symmetric encryption standard, like AES symmetric, and then registers the encrypted contents — the “cipherbook” — with a storage-centric, decentralized filesystem like IPFS. Whenever the publisher so chooses to sell this book to a customer, they transact and in doing so ask for the customer’s public key, using it to encrypt the cipherbook’s symmetric key and safely embed it as some compiled function* that gets written to a blockchain. This function might take as input a customer’s private key, the IPFS address of a cipherbook and perhaps additional details like the key of an approved vendor app, as well as a scope of content to access and returns the scoped portion of the book to the app. This compiled function serves as both a receipt and an access mechanism for the customer who has purchased the e-book, ensuring access even if the publisher behind the product ceases to exist.
*Author’s note: The logistics of storing a compiled function to a decentralized blockchain needs further exploration and is not my specialty. Perhaps a key management system (KMS) like NuCypher may be an interesting case study.
In order to actualize the customer’s first sale doctrine rights, a smart contract system could be used to stipulate temporary transfers of access between customers, with the smart contract automatically returning ownership to the original owner after the lease period expires. This same mechanism could be used to support re-sale or loans. This would enable customers to resell their purchases or loan them to others without the risk of the contents not being returned, while also providing publishers with a safeguard against mass duplication.
Overall, a Secure Model for Ebook Ownership, combined with blockchain technology and smart contracts, could provide a secure and customer-friendly e-book sales model that benefits both publishers and customers. Publishers would have greater control over the distribution of their content, while customers would have ownership rights and more flexible access to their purchases. Libraries also would be empowered to participate in long-term preservation and equitable collection building of born-digital materials. By collaborating on the creation of a shared infrastructure for e-book sales and distribution, publishers and technology companies could create a sustainable and equitable ecosystem for the sale and distribution of e-books for the next generation of e-commerce.
