March 29, 2004

Why News and Technical DRM Don't Mix: Linking and Linking Expression are Key

Nonfiction media, the news, has 24 hours of high value, and then the content turns into something else. It's a swift transition news content makes into the status of history, an archive of facts and information, a social accounting, that has a different meaning and use after it stops being news. The production of it relies on various institutions and practices, as well as political, social and economic forces that change over time and shift it's place in society. The press is privileged, as it has constitutional protections for freedom of the press, libel protections, and other legal protections, not to mention access to governments, records and situations that the rest of the public doesn't have. There is a reason why we don't force the press to be licensed or pay a tax, and grant some level of immunity for search and subpoena or prior restraint. These privileges are there in exchange for certain responsibilities the press accepts in a kind of social agreement with users to serve the public with useful, trustworthy information, or find itself irrelevant and disdained.

At the Mediamorphosis (API) conference two weeks ago, one of the three workgroups of participants proposed using Digital Rights Management tools (DRM) to help insure secure content. These weren't technologists but I don't think they meant a firewall to make their content secure in archives, but rather DRM that would prevent the content from being opened by anyone anywhere outside their firewall that they didn't authorize. The subject then came up in the blog where it was noted that Michael Silberman (MSNBC) said: I think DRM could be used to keep people from stealing, and get them to pay for content. And it could be used to facilitate the making of content.

I think some of this desire for DRM on the part of creators of news content has to do with thinking that when users access free news content online, it means to creators that users don't value the content because they don't pay for it. Under the old model, paying for a newspaper meant to users and creators alike, that they were paying for the paper and the delivery, and to the creators, that their content was valued because it was bought. The newspaper's business department knows the real story, which is that those subscriptions don't cover the cost of content, ads and classifieds do. Subscriber zipcodes helped sell the ads so subscriptions attracted revenues and defrayed the cost of paper and delivery. But perhaps they also said something more about the need by those who report the news to feel like their work is valued. Corresponding with this is the desire to protect what they see as the "free" accessing of their sweat and hard work, with little regard by the user for that work.

What some reporters don't see is that online the content's value is expressed through users linking, thereby expresssing their own attention as well as referring other users, under an ad model, or clicking through RSS feeds to the content websites as well as the general authority generated by past good work. Users value good work and they show it by coming back and by linking or following other's links. (I understand that the current biz model relies heavily on paper business models paying for content generation, followed by the repurposing of that content under a licensing fee to their websites that use the ads to generate revenue. But I do think that eventually content creators will figure out how to leverage ads to pay for that content generation online.)

DRM the way Silberman describes using it is something each content maker would define, where they would decide what sort of restrictions to make on users accessing and distributing the content. Ordinary users, if they have trouble opening the article, sending it to friends and family or saving it indefinitely, all of which is annoying, will abandon the information because using it doesn't reflect the social norms they understand with fair usage of news content, and it confuses them, but hackers will figure out how to get around it. And imagine the chaos for users as they access different content makers' work, each with different settings and restrictions.

DRM is a different technology than a firewall in that DRM is wrapped around the media and goes to the user's machine, whereas the firewall resides at the servers of the content provider and is a barrier to entry. DRM as both the technology solution, as well as a legal structure, is not a sound way to go with this content most valuable for 24 hours, versus say a work valuable for many, many years, something like a novel or movie or music might constitute, which is a different issue and argument for the social, copyright and technical issues. News content has the potential, if you share it, to keep you in front of users as an authority and make users happy to be a part of your information community. Users will go elsewhere if they meet technical difficulty, and if the information is not available under less complicated circumstances, they will abandon their search for that content in favor of some other topic or content they can get more easily. But this is a use and technology argument against DRM for news and I don't think that is the most important issue here, though it is important.

There is also the issue that there has never been DRM that has not been cracked eventually. That being the case, technical news DRM would eventually be cracked by those who want access (as opposed to legal DRM which might make cracking illegal, like the DMCA, which I don't want to get into here, but it has been discussed quite a bit in other posts if you want more information.) Practically speaking, it is unlikely that news DRM would work any better to achieve the goals of the news makers, than it has for record companies, movie makers or gaming companies. But this is also not the most important reason not to implement DRM in news content.

There is a point to consider in the case Ernie Miller wrote up a while ago about the copyright case on newspaper headlines in Japan. I think though that something similar here in the US would not win because the title of an article would fall under the "names, titles and short phrases" that don't get copyright protection, partly because they are factual, even if they are a kind of expression (tends to fall more in the trademark area of IP for names and phrases). Therefore, using DRM to completely restrict an article, to the extent that it denies the user access to the title, author, publisher and date as unprotected metadata, might keep users from seeing this metadata. This metadata is also content in a way, and contains some factual information around the event that isn't really copyrightable at all. Also, the amount of effort needed to create and publish the work significantly affects the value of the work and the way the law treats using bits of it under fair use, even if our copyright laws are too obtuse and out of date to recognize these as well as the specific digital vs. analog media differences. But users instinctively see this, as the use the media, and therefore expect different things from digital news media than from analog articles, expect to share articles they see with people in email, quote from them in blogs and repost headlines; they have more expectations with digital news media.

The most important reasons news media companies and creators should not implement DRM is because of fair use considerations of the content itself, as well as the maintenance of their positions as reporters of news, and authorities of information.

Online, bloggers and other web content makers use and depend on traditional journalism by discussing news within their writing, as well as by linking, making traditional journalism a kind of authority. These users are filtering for audiences, pointing to things, saying to their readers: look at this for some reason, and here, I'm telling you what I think about it, and here's the link to the article itself, to some other backup to do with the subject, to someone else talking about the same thing. Snippets of content, used because of fair use, commented and fisked, are key to this as well, to show what is being discussed. Imagine Roger Ebert having to review a movie without the clips, describing the whole thing. It's possible, but not nearly as powerful as being able to cut and paste something that needs to be shown.

So Reuters announces plans to use FAST ESP or Fast Search to scan for copyright violations across thousands of feeds. Tom Curley at AP announced last Fall that he wanted to wrap AP content in DRM. Both are misguided attempts to control their business models as they are disintermediated by digital media (and like every other industry facing the paradigm shift due to the information age, it means sorting out a new business model and changing, not holding on to what you've got -- or you'll find yourself in the company of buggy producers). It's not that they shouldn't police unauthorized commercial use of their products. But for anything but those problems which could be solved by search systems like FastESP, they should abandon DRM and other technical self-help methods to keep people from their content. Using any technology that gets in the way of users interacting with the content annoys people and lessens their overly-informed and highly mobile audiences. This is not a way to win friends or be an authority. And, if a company is using a Reuters or AP feed in an unauthorized manner, it would seem to me they could contact them directly. If the corporate user wasn't willing to comply, and are outside of any legal bounds, what is to stop that same illicit feed user from just hacking the DRM? In the end, those who want the feed will hack it, and all the folks who are just readers with a lot of choices will move on to other sources when they can't open or talk about the content online.

Here's what you want: users who, every time something in the world happens, think, hey, I need this media company's content, this writer, this site's take. If you use DRM or make barriers, you will reduce your standing as an authority for news both as content and as linking expression because invariably some won't be able to open it or link to it. If you make yourself unlinkable, you will cause yourself to be irrelevant across the influencers on the internet that point to the sources, filter them, for other users. Who links to the Wall Street Journal? In Technorati, they have 354 links compared with the NYTimes at 39,412 and the Washington Post at 21,319. Who do you think has more authority online? The paper with premiere content in its niche and 600,000 online subscribers, and a lovely firewall? Or the paper of record. Now imagine losing that authority with the DRM you wrap around your articles.

You're nothing online if you're not linkable.

Posted by Mary Hodder at March 29, 2004 09:14 AM | TrackBack

although the Washington Post's numbers should start to go down now that they require registration.

Posted by: joe at March 29, 2004 11:43 AM

The newspapers CAN have their cake and eat it too. But they've got their pricing models backwards. New news should require a subscription BUT the links are permanent and will work for everyone (subscribers and non-subscribers) after three days. Subscribers won't pass around their access codes because it will affect their reputation. For more info please go to


Posted by: Cynthia Typaldos at April 1, 2004 03:20 PM

The NY Times also makes you pay to access articles after 7 days of it being published I believe.

Posted by: anon at April 2, 2004 03:51 AM

NYT and many others do make you pay after 7 or 14 days, but for the time those stories are most important, they are free and linkable. Also, there are ways to get at the stories after they go behind the firewall. For example, the WDCPost, if linked, will maintain the link beyond the 14 days content is searchable. After that, if you search for something 15 days old, they charge. But if we link in a post, it remains available directly.

Posted by: mary hodder at April 2, 2004 07:39 AM


I read your excellent, thought-provoking piece on DRM'ing the news. (I'm a bit confused on what blog it is on, so wasn't sure where to put these coments, so am sending them directly to you. Feel free to post them to the appropriate blog if you care to.)

I would further emphasize the "hassle-factor" of DRM, particularly if used to encrypt individual articles. There are technical hassles (installation, use problems, multiple versions, obsolete technology/vendors), sharing/saving hassles, administration/licensing hassles (who handles disputes efficiently?), as well as the linking issues you mention.

I would also agree that history shows us, if the content is considered to be valuable enough, hackers will cut through the protections. If not valuable enough to encourage hacking, what's the point?

The peculiar time-dependency of news' value is also important. You say that it has "high" value the first 24 hours. Well, for the first 24 (or fewer) hours it is "news", but, given that it is largely considered to be a commodity and is almost always provided at no charge, I'd question the "high" value. But, in any case, there is some value during the first 24 hours. There is another value after it is in the archives and retrievable through a paid mechanism. And there is a third value during the period after the first 24 hours (when it is no longer current "news") and before it qualifies for paid retrieval.

Even if the value was sufficiently high to justify protection, it is hard to imagine a DRM system that could handle such time-based variability efficiently.


Terry Steichen

Posted by: Terry Steichen at April 13, 2004 09:25 AM