Loser: DVD Copy Protection, Take 2
The next generation of technology won't have a Hollywood ending
In 1999, a 16-year-old Norwegian high school student took on the motion-picture industry and won. The teenager, Jon Johansen, wrote software that decrypted the Content Scrambling System (CSS) that rearranges the bits on prerecorded DVDs to prevent the discs' being played back on unauthorized hardware. Until Johansen wrote his software utility, which he called DeCSS, you could copy the bits from a DVD to your computer hard drive, but because those bits were scrambled, you couldn't play a movie from those copied bits.
Johansen's DeCSS soon reached every corner of the globe, thanks to the Internet, and prompted a host of legal battles. In the end, Norwegian judges ruled his actions legal, but U.S. judges ruled the distribution of the software illegal. Nevertheless, anybody who wants to make a copy of a DVD can now do so easily, much to the dismay of the entertainment industry. It is not clear whether such copying is done legally under copyright law, but no individuals have been prosecuted for making copies for their own use.
With manufacturers about to unveil a new generation of DVD players and discs, moviemakers now see a rare opportunity to get the horse back into the barn and lock the door tight. So, this past July, two entertainment companies joined with six electronics manufacturers and chip makers to announce the creation of the Advanced Access Content System (AACS), the copy protection scheme designed to keep future generations safe from pirated DVDs. The specification was due by year-end 2004, and products incorporating it are slated to appear by year-end 2005.
Backers of the protection method are betting that AACS technology will finally thwart unauthorized copying of DVDs while allowing consumers to distribute movies legitimately over networks within their homes, play them on a variety of devices (standard televisions, portable movie players, and laptop computers), and store them on home media servers. "We wouldn't be investing our time otherwise," says Michael Ripley, the chairman of the AACS alliance's technical working group.
But critics of the technology say it is bound to fail in achieving its most important objective—blocking wholesale pirating of DVDs—and it may irritate consumers if the promised in-home distribution isn't quickly forthcoming and easy to use. The AACS project "doesn't make very much sense," says Seth Schoen, staff technologist at the Electronic Frontier Foundation in San Francisco. "The commercial copyright infringers in Southeast Asia that burn billions of counterfeit discs will not be deterred by this." AACS
Goal: To develop a specification for copy protection of next-generation DVDs why it's a loser: It won't deter commercial copyright infringers
Organization: AACS License Administrator
Number of people: Not available
Budget: Not specified
Efforts that bring together industries with competing interests are challenging at the least and have often broken down acrimoniously. Witness the failure of the Secure Digital Music Initiative, an attempt to bring watermarking technology to digital music. Also keep in mind the different goals the entertainment and electronics industries certainly have. Movie moguls would prefer to make all copying impossible. You want two copies of a movie? Buy two. By contrast, electronics firms would prefer relatively unrestricted copying, because being able to move videos you already own to new devices (such as media servers or handheld players) helps drive sales of those devices.
"Any of these [electronics and entertainment] consortiums runs the risk of collapsing under its own weight," Michael McGuire, research director of the media team for GartnerG2, in San Jose, Calif., told IEEE Spectrum.
The AACS group was founded by IBM, Intel,Microsoft, Panasonic (Matsushita Electric), Sony, Toshiba, Disney, and Warner Bros. Studios. The alliance is optimistic, says Panasonic representative Richard Doherty. "So far, it's going very well, though discussions are spirited at times," he adds.
"We are confident that AACS technology will be part of the next generation of content on optical media," declares AACS's Ripley, who is a senior staff engineer at Intel Corp., in Santa Clara, Calif.
The key to the spirit of compromise is an agreement that the AACS specification will allow consumers to move the data on an optical disc to the various devices they own, including video servers and portable video players, either directly or via a home network. In all the scenarios developed by the AACS alliance, that data would exist on the disc in encrypted form. It would stay encrypted when transferred to other devices and would be decrypted by those devices. The details of this portability have not been announced, but the technological underpinnings are expected to be included when the first version of the copy protection specification is released.
Once products do ship, potential crackers—including teens in basements, academics, and large-scale pirates—will undoubtedly tackle this new encryption challenge. Some details of the specification were apparently still under discussion as this article went to press, including where on the disc the encryption codes will be physically stored.
But one key parameter has been made public. The CSS encryption in the first generation of DVDs, which Johansen defeated, used a proprietary 40-bit key for encryption. AACS will use a so-called strong key, the 128-bit Advanced Encryption Standard approved by the U.S. National Institute of Standards and Technology.
"This removes one of the obviously dumb things from CSS—they were using a cipher that was easy to break," says Dan Wallach, assistant professor of computer science at Rice University, in Houston.
Because even strong keys can be compromised, the heart of the new protection technology will be its ability to keep on protecting data even after it has been cracked.
The basic idea in recovering from cracking is to make a compromised player key obsolete. Compromised players could continue to play old discs, but not new releases. And crackers would have to start all over again.
Ripley identifies a technology called media key block as an important element in recovering from cracking. With this system, there are actually two keys—one is on the disc itself, but it doesn't work until it is decoded by a second key installed in each player. Multiple versions of this second key can exist; indeed, it is possible that each player would have a unique key, or that groups of players would share keys. Either way, if one key is compromised in the way that CSS was compromised by Johansen and if that decoding method becomes public, new DVDs could include updated on-disc keys that would cause the compromised player-based key to fail. They would, nevertheless, still work with other, uncompromised player-based keys.
Under this scenario, it seems that additional methods might still be needed, since someone who cracks a key but doesn't broadcast that fact could fly under the radar and be able to continue to copy new releases.
AACS's plan shows that the industry has learned from its embarrassing experience with the Norwegian teenager. But that lesson doesn't solve the whole problem, says Rice University's Wallach. "When you give a secret to a million people, and one of them reverse-engineers it and releases a movie, you might not even be able to identify what key you need to disable," he points out.
GartnerG2's McGuire says he appreciates the industry's need to address cracking, but he also sees the cycle of crack/fix/crack as an endless loop. The motivation to crack high-value content like movies isn't going to go away, he notes, so the overall economic benefit to the industry is not obvious.
"The movie-studio people are thinking that by doing this, they can maybe slow commercial copyright infringers down for two weeks, and that would improve their bottom line," says the Electronic Frontier Foundation's Schoen. "I would be impressed if that is the case. No technical measure ever has been proven to have an effect on these people."
Perhaps the biggest question of all is: Will consumers who have become used to freely copying DVDs for viewing on their hard drives or for sharing with friends accept a new technology that restricts such copying?
"We can understand that the technology companies and movie studios have become attached to the idea of creating proprietary encryption for optical discs, but it doesn't do much good in terms of the stated objective of preventing copyright infringement, and it has a lot of downside for consumers," says Schoen.
He and others say success or failure of the approach may hinge on whether consumers will be able to move videos around a home network and among multiple players right from the outset. If they can, they might be more willing to accept the new technology. "It might lead to a higher level of satisfaction for mainstream consumers," Schoen says. "So there will be less pressure from them for unauthorized products that allow backups. But at this point, the details are a bit sparse."
If they can't, individual consumers will find ways to transfer their content anyway, Wallach says. "It is not a matter of if—it is a matter of when. As long as I have the technology in my living room to watch it for myself, I can modify the system to extract the video. They can make it hard, but they can't make it impossible.
"They are living in a fantasy world," he concludes.
To Probe Further
For official updates of the progress of the AACS specification, see http://www.aacsla.com. Information about the Electronic Frontier Foundation is available at http://www.eff.org.