The RIAA's Low Watermark

SPECIAL REPORT The SDMI Challenge produced scientific evidence that encryption can’t stop piracy. So the industry tried a new tool to secure digital music: censorship. There’s no such thing as a pirate-proof distribution system, according to a peer-reviewed scientific paper by Princeton professor Ed Felten and eight other researchers. But the RIAA aims to make […]

SPECIAL REPORT

The SDMI Challenge produced scientific evidence that encryption can't stop piracy. So the industry tried a new tool to secure digital music: censorship.

There's no such thing as a pirate-proof distribution system, according to a peer-reviewed scientific paper by Princeton professor Ed Felten and eight other researchers. But the RIAA aims to make sure you read no such thing.

During three weeks last fall, Felten & Co. cracked the SDMI Challenge posed by the Secure Digital Music Initiative consortium, deciphering the contest's four watermarks and finding bugs in its two other tests. "There was a lot of controversy between us and SDMI about whether we had in fact done that," he says. "Our position was: Wait for our paper - the details are in there." But when Felten - known for his work on Net security and his testimony for the DOJ in the Microsoft antitrust trial - went to present his findings at the Information Hiding Workshop in Pittsburgh this spring, the Recording Industry Association of America and SDMI raised the threat they might sue.

"The RIAA tried to block the paper, I think because it is evidence that our claims are correct," explains Felten, who withdrew the paper. Though it showed up on the Net days before the conference, Felten's felt a chilling effect: "We've been unable to publish the paper through normal scientific means and unable to talk to our colleagues about it at conferences, where 100 or 200 researchers are all in the same room."

Felten says it's important to note that his findings weren't based on eureka moments. "There wasn't any big breakthrough," he says. "If any of these technologies were deployed widely, they would be broken fairly quickly." Watermarking presumes that the inner workings of a system remain secret - a mistake Dutch cryptographer Auguste Kerckhoffs first warned against in 1883. "Anything that is widely distributed in an economical device will be subject to reverse-engineering," says Felten. "You can't call something a secret when you expect to deploy it into the homes of millions and millions of people."

MUST READ

The RIAA's Low Watermark
Felten Talks Through the Paper
People
Jargon Watch
Bandwidth Blaster
Calculated Communiqués
Friendlier Fire
Anti-Drowning System
Wired Index
Raw Data