Copyright, privacy right and usageright


In this post I return to a topic upon which I have commented before. That topic is intellectual property rights for digital media. Today, however, I hope to express myself more clearly and give a better description of both the problems and my proposed solutions.


Before describing the problems with our current intellectual property law I think it would be helpful to set out the various relevent facts and useful principles. Firstly we will look at the properties of intellectual property as compared with physical property. Secondly the differences between analogue and digital storage technologies. Thirdly facts concerning distribution networks. Fourthly principles of policing and fithly the ways in which encryption is used and abused. Only then will we have enough of a foundation to proceed.


The properties of intellectual property that physical property lacks:


A piece of literature is more valuable the more people can read it (because more people can enjoy it). A time saving idea is more valuable in a community of six billion than one thousand. Intellectual property (IP) can survive even if foreign nations do not respect it but the value of IP increases if they do. Many people can possess the same piece of IP at the same time and I can aquire a piece of IP without taking it away from anyone.


The properties of digital IP that anologue IP lacks:


Digital IP is very easy to copy (and the price of copying is currently following a version of Moore's law). Digital IP can be modified easily. It can be encrypted and decrypted quickly and efficiently (more so with time). Digital IP is easy to detect when it is not encrypted. Finally a piece of digital IP is not tied down to any one physical medium. It is very close to the theological concept of a soul in fact!


Facts concerning distribution of IP:


For a piece of IP to be displayed or used in some way an unencrypted version of that IP must exist at some point. On the other hand if a piece of IP is merely being stored there is encryption that is believed to be uncrackable that can hide the piece of IP from view. The strength of this encryption depends on our continued failure to construct quantum computers and on a widely believed but unproven mathematical conjecture. Finally when distributing data to multiple locations simulateously the most efficient method is often to make many local copies of the data and then when anyone requires access to the data for it to be sent/copied from the closest local copy. For popular IP the system can be more than an order of magnitude faster than a competing system which minimises the copies made.


Principles of effective policing (and rationals for IP law):


Detection is often much easier than prevention. Anyone may suddenly decide to commit a crime. Police cannot be expected to prevent a previously law abiding citizen from commiting an offense (indeed the expectation that they do is a major headache in counter terrorism). A much better strategy is to try to detect crime and identify the purpetrators. Laws generally require the backing of the community. This is because (without draconian measures) the community is necessary to help identify criminals and provide evidence. Intellectual property law was originally created in order to provide an incentive for people to spend their energy creating things. It works by allowing people to claim back economic benefits from those that make copies of their creations. Intellectual property law also allows people to control the usage of their private data. This aspect of IP law really protects people against the abuse of their personal data. Finally the punishments for offences need to be in proportion to the severity of the offence (part of the problem with the RIAA's series of lawsuits. Our courts, the prison system and our police force are rendered less efficient by laws which are not proportional. Police forces & the courts are asked to waste time on trivial offenses and the prison system requires extra capacity to accomodate small time offenders. Finally the less proportional a law is the more tempting it is to abuse it. Diebold attempted to use the digital millenium copyright act to silence criticism of the appalling lack of security in its e-voting machines.


Facts about the use and abuse of encryption:


There are many civilian uses for encryption incluing protecting sensitive private data, protecting corporate secrets, completing financial transactions, proving ones identity, protecting ones identity and engaging in various cryptographic protocols (online gaming, secure voting etc.). However, criminal uses also exist including hiding fraud, money laundering, exchanging dangerous information including terrorist training, the genomes of deadly viruses and atomic secrets. But also included in this list at present is copyright infringement. This is a problem because copyright violations are not nearly as serious as other items on the list. See this blog post of mine. For the other items on the list we may need to empower law enforcement and the secret services to require access to encrypted data (backed up by legal sanctions). However, these measures cannot be used to prosecute copyright violations because such a measure would not be proportional and would undermine community support for police action against the more serious crimes listed above.


Now I have laid down the foundations I can proceed to build my argument. Firstly I shall investigate if copyright aplied in the digital realm achieves its objectives. Secondly I shall list ways in which we might try to control copyright infringement and the fatal problems involved. Finally I shall propose an alternative to copyright and explain why I think it is suited to the problem.


Does copyright accurately protect the economic incentive to create (in the digital realm)?


A digital copy of a piece of IP does not necessarily have any value. A copy of a book that just sits there is not helping the economy or the owner in any way. Of course knowing that you can access a piece of IP is valuable but this type of value is really created by the library or storage device not the piece of IP itself. However, using a piece of IP in a calculation can give an economic benefit. Displaying IP can create value (on the screen, sound system or other display format). Notice that both of these usages require the copyrighted work to appear in an unencrypted format. It is simply impossible to use IP without (at least temporarily) having an unencrypted version of it. Finally as we have already mentioned efficient distribution of digital data requires the creation of many copies which are themselves only copied but not never used.


Does copyright accurately protect privacy?


Firstly we should note that a piece of private data can only be abused if it is found in an unencrypted format. The more pieces of private data that exist in one place the worse the potential abuse may be (see my post on the matter). However, copyright laws fail to distinguish between encrypted data and plain text. Note that with encrypted data it would not be possible to even detect the presence of the private data.


How might companies and the government attempt to control copyrights abuses?


1) Attempt to prevent people making unauthorized copies. (Digital rights management, protected DVDs etc.).


This is an extremely foolish way to proceed for two reasons. Firstly it attempts to prevent a violation of copyright law rather than detecting it. Secondly whenever IP is used there must exist a plain text version of it and it is impossible to prevent people from copying this (outside of nightmarish totalitarian regimes).


2) Target distribution networks. (lawsuits against Napster, pirate bay etc.)


This is not going to work primarily because the best distribution networks are distributed, they are useful for scientific and other legal civilian purposes (such as user generated content). The distribution networks cannot be expected to police their data becuase the data may be encrypted.


3) Detect origin of copyright breaches (Apple appears to be using this with low key water marking systems).


This is a considerably more sensible idea than (1) or (2) but there will still be too many copies for the system to be practical (unless the data is all centrally stored which is itself uneconomic). The wrong people will often end up being punished just because someone else copied their data without their permission. Finally water marking is not practical for all types of data.


All the methods fail to allow useful economic activity by punishing the creation of huge numbers of copies that do not result in any economic loss to the creator. As a result of this any method that attempts to enforce current copyright law will result in a community that will not support the law.


I will now propose my own solution to these problems. It is a radical idea but I hope to explain why it is a better idea than our current law.


I propose that for digital data we abolish copyright and replace it with a usage right and a privacy right.


The usage right would grant the creator the sole (but transferable) right to display the IP or to use the IP in a calculation. The privacy right would grant the owner (of private/confidential data) the sole right (also transferable but possibly with caveats) to have this data in an unencypted form.


The usage right makes sense because it possible to detect usage as the data must always occur in an unencrypted form. To enforce this law it is only necessary that violations of usage right can sometimes be detected. The police or alternative enforcement agencies can then use that possibility to discourage abuse. There are many ways in which usage rights might be enforced and there should be discussion as to which should be chosen. Usage rights are compatible with efficient methods of distribution. Usage rights lower transaction barriers to creating valuable modified forms. Modified IP which is similar enough to the original will also be similar enough to be detectable in its unencrypted form.


The privacy right makes sense because it isolates the cases where abuse is likely, most serious and detectible whilst ignoring instances where abuse is difficult but impossible to detect.


Note that with usage right it would be possible (even desirable) for payment to work on either blanket licensing or on a form of metering (I would preper a stratified combination).


Finally we should address the issue of community support. It should be acknoledged that there will always be a black market for IP. But in the presence of efficient and accurately targeted means of detection black markets must operate in secrecy and without high bandwidth connections to the rest of the Internet. This will mean extra costs for being part of the black market. To ensure the majority of people join the legal market it must simply be the case that the benefits of joining outweigh the costs of paying for usagerights. I argue that this wouldn't be too hard given the benefits that would come from usage rights.


The benefits of being heavily connected to the Internet are great and growing. Usage rights will dramatically reduce transaction costs and new services will become common (think recomendation, links, enhanced content etc.). Virtual libraries will become common with

some IP linked to and some stored locally. New types of media will be created by mixing and linking data together with no walls or barriers from traditional copyright. New versions of IP (song title and band information) will be available for automatic download. Finally with the freedom to browse the sum total of human creations (minus private data) and pay only for what is actually used a new, more vibrant and connected market will encourage people to be even more ingenious, to remix, to link and to design their creations to be remixed and linked in with existing culture.


Note that this is a page in progress and that I may come back to edit it to hone my arguement.