Why They'll Need Open Standards
Many studios and record labels have attempted or are attempting to start Internet-based businesses. Napster was arguably the most successful Internet entertainment venture to date. Many consumers want to use a service to share music on the Internet, but today there are likely few services that are operating completely within regional laws, at least in the U.S. Rather than serve up compelling businesses such as Napster, some big entertainment companies have served up the Hollings Bill, which places severe constraints and costs on digital devices by requiring CPS to be embedded in digital devices. (See the sidebar, "Race Against Time.")
The problem with forcing CPS technology is that copyright protection systems cannot prevent copyright violation, as repeatedly demonstrated by DeCSS and other tools. Even if someone were to invent an unbreakable CPS that was widely deployed for, say, movies, a customer could nonetheless rip a movie after the signal has been decrypted and put it onto the Internet for widespread distribution.
Conventional law enforcement will be the only recourse for rights holders. Complex CPS technology behaves as Schneier described it, it doesn't deter the professional and is overkill for ordinary users who might be tempted to make a copy.
Fortunately, selling copyrighted content without tight restrictions isn't always considered sinful. A compelling content-trading service will let its customers use content works in a variety of ways. If a consumer wants to make a copy, the company will provide a service to make copies; if the consumer wants to write the content work to a DVD, a good service will provide a DVD writing service; and so on. A technology VP from a major media company told me that his firm wasn't opposed to copying per se, but that the company doesn't yet know how to match consumer preferences for using content works.
This is not an engineering problem, but a series of legal and logistical issues among rights holders. Even a copyright protection system is secure when its users and owners have a stake in protecting its secrets and keeping it functioning properly. It's likely that most people prefer to spend time enjoying content over configuring nodes on an illegal content-trading network.
Because we don't conclusively know consumer preferences for using Internet content-trading, and because the necessary business models aren't in place, technologies must be very flexible. Open standards are flexible because they introduce modularity and interoperability. Internet open-standard technologies may not hide content works on computers or enforce copyright provisions in an entertainment appliance, but open protocols provide the infrastructure for businesses that trade in content works on the Internet. These businesses need open development.
Where Will It End?
This is less of a rant against the DMCA and more of a prediction. One might reasonably predict that the DMCA will encounter stiff resistance in the short run and leave little effect on open standards in the long run. This is because the underlying content protection technologies are inevitably weak, even though they may use some of the most important technologies in the public domain (such as the Advanced Encryption Standard or Internet key management). It's hard for me to imagine that U.S. and international research and development into such technologies could be shut down by the DMCA. It's certainly possible, but hopefully the impending legal battles will loosen restrictions.
Open standards guided by the principle of end-to-end interoperability provide a cheap way to interconnect devices on a global scale. It's cheaper for the provider and the user because open standards make it possible to remove a part without drastically affecting the whole. If a particular CPS or player or codec isn't suitable for a service, it's easier to replace when these devices connect through standard protocols. This is a cost savings to operators and their customers; it also encourages rapid deployment of new devices.
Standards built on licensed technology complement open standards when licensing terms and conditions are needed. A licensing authority can constrain a licensee to use a technology in secure or safe ways. Even the best technology, such as the RSA algorithm, can be compromised when used incorrectly. Big entertainment and consumer-electronics companies will likely insist upon using licensed technology for their information products, and will want technical protection measures to protect copyrighted products.
On the Internet, however, such copyright protection systems will need to use standard services for naming and identifying digital items, for locating digital items, for secure transactions with Web sites, for encryption, decryption, and key management. The infrastructure necessary for digital rights management and copyright protection systems are only a small part of it.
It's unreasonable to expect that all of these technologies and standards will be duplicated in the form of licensed technologies. It's more likely that licensed technologies will use open-standard technologies that must continue to be developed without hindrance from the DMCA.
Mark Baugher is co-chair of the Internet Research Task Force DRM Research Group and former chair of the Internet Streaming Media Alliance DRM Task Force. He contributes to the GDOI group key management protocol in the IETF MSEC Working Group and the Secure Real Time Transport Protocol in the IETF AVT Working Group. Mark joined Cisco Systems in 2001 and currently works on cryptographic security in the Core IP Engineering Division.
