Features

net.wars: Distrusted systems

The news this week that the European Commission is investigating Microsoft's and Time-Warner's April acquisition of ContentGuard is not trivial, and it's not just another case of "Oh, let's Get Microsoft". What few outside of the digital rights management arena have realised yet is that ContentGuard holds very significant patents that could have a very broad impact if the use of digital rights management takes off.

The back story begins with Mark Stefik, who was and is a researcher at Xerox PARC. Stefik is a lovely guy, but in the early 1990s he came up with an idea that librarians, public domain advocates, and information-wants-to-be-free types welcomed with all the enthusiasm of Greenpeace contemplating a nuclear bomb: trusted systems. In his seminal essay, "Letting Loose the Light", Stefik proposed the notion of a repository which would store documents and know what rights were associated with them. Printers, computers, other devices, might have such repositories; when you tried to access a given document the repository would know which rights you had paid for, and these could be quite fine-grained.

It would be possible to specify reading, printing, display, copying, backup, and other types of rights, and the system would be able to charge you appropriately. Stefik's idea was that a lot of content was not being placed on the Internet because people were afraid it would be massively copied, and this type of system would encourage content owners to let loose the light - that is, distribute their content without fear.

These ideas were discussed at the 1997 Computers, Freedom, and Privacy conference - unfortunately, the only one from which no papers or recordings are generally available. Stefik presented some of his ideas in a panel that included Eugene Volokh, Pamela Samuelson, UK barrister Alistair Kelman, and Georgetown intellectual property law professor Julie Cohen. Samuelson compared trusted systems to walls and talked about the then recent WIPO; Kelman quoted Thomas Babington Macaulay's description of copyright to a "tax on the innocent pleasure of reading", and Volokh risked death by saying that contracts were the way to do these things and if you paid for public domain content because some electronic publisher had wrapped it in a charging mechanism, well, that was just "bad shopping". Cohen was the most interesting: she talked about the privacy consequences of having systems that logged (for billing purposes) every access.

Stefik and Xerox began filing patents on his ideas in about 1994. In 1998, Xerox and Microsoft teamed up to launch the fruits of Stefik's ideas as ContentGuard; the launch, held at PARC, featured presentations from Steve Ballmer and various Microsoft folk showing off how comfortable it was to read ebooks on Pocket PCs. By then, Stefik had moved on to other things to such a degree that the PR person arranging things didn't even know his name. The patents that he and Xerox filed now, of course, belong to ContentGuard

Skim the patents and you'll see they're impressive. They cover using markup languages to attach machine-readable rights to content; they cover incorporating charging mechanisms. In fact, they cover much of what we'd describe as digital rights management.

The published information about the EC's investigation suggests it's most concerned about Microsoft's dominance in the operating systems market. The EC is right to be concerned about this, and it's right to be doing it now. In the past, anti-trust investigations have been five years out of date. They were prosecuting the browser wars when they should have moved on to media players, and it's not clear whether they've gotten to media players soon enough. In this case, they may be acting at the right time. It is unquestionably a risk for the future if Microsoft can use its operating system to dictate the choice of digital rights management systems and become a single gateway through which consumers must pass to access everything from books to movies.

But the most significant piece of this is the patents. ContentGuard's Web page states that its patents are fundamental, and seems to believe these patents cover any rights language. This would mean that someone wanting to create an open-source alternative that served non-commercial values might not be able to do so without negotiating with ContentGuard. If the other half of ContentGuard is owned by Time-Warner -

Unlike many software patents, I've seen no criticisms that these are over broad or awarded for ideas that are not novel. To be sure, markup languages have existed since the 1960s, but they were being used to determine how content was displayed, not to control how it was accessed or to charge for it. A court case could settle it - at a cost of millions. It would certainly be unfortunate if the only argument against the validity of these patents were that their social costs are too great to let them remain in force until they begin to expire in about 2014. The GNU licence may perceive proprietary rights as damage and route around it, but these patents may enable two of the most dominant corporations in the world to do its own rerouting. The EC is right to be worried.