Losing the war on patents

Attempts to fix the intellectual property system from below are faltering. Is it time to bring in the feds?

Published February 15, 2002 8:30PM (EST)

Just three months after it launched in the fall of 2000, BountyQuest.com, a self-described "patent-busting" Web site, declared itself a success. In January 2001 it awarded four $10,000 prizes to individuals who had presented evidence that BountyQuest believed would successfully undermine a set of specific patent claims.

One award attracted widespread attention. In April 2000 a company called InTouch sued Amazon.com and five other companies for infringing on a pair of its patents covering a type of music sampling -- technology that lets users preview part of a song and tracks their preferences. But less than a year later, BountyQuest announced that it had found evidence of "prior art": an example of a similar digital music sampling technology that long predated -- and thus undermined -- InTouch's patent.

The Amazon angle added extra intrigue. Amazon, at the time, was the target of vocal criticism for its own heavy-handed approach to seeking patents on everything under the Internet sun. And Amazon's founder, Jeff Bezos, was an investor in BountyQuest. So when one of BountyQuest's first forays into patent-busting turned out to be a potential boon for Amazon, some eyebrows were raised at what could easily be seen as a conflict of interest.

Amazon itself wasn't the only problem with the patent system, however. To many observers, the corporate rush to stake intellectual property claims on every possible aspect of Internet life threatened to strangle future creativity and innovation -- not to mention trample all over work that had been done years earlier by programmers who had donated their code to the general public.

BountyQuest, as envisioned by its founders, was going to be a market-based approach to patent reform. By providing a forum in which interested parties could offer rewards for the discovery of patent-busting prior art, BountyQuest aimed to solve the problem of an out-of-control patent system.

One year later, BountyQuest's supposed patent-busting experiment appears to have itself gone bust. Only about 20 BountyQuest contests have resulted in bounty-winning prior art since the site's inception. New bounty offers, meanwhile, have slowed to a trickle; 23 contests were running when the company launched but only three contests are now open for submissions.

Most troubling of all, the InTouch case suggests that even in instances where BountyQuest has generated results, the company still hasn't achieved much. Last month, two defendants in the InTouch lawsuit, Amazon and Liquid Audio, settled their patent infringement cases with InTouch, agreeing to license the InTouch patent rather than fight it. And in both cases, say parties involved in the settlement, BountyQuest's discovery never came up.

"It was never used in the case," says Joshua Kaplan, founder of InTouch, a 12-year-old digital music company based in Berkeley, Calif. "The defendants didn't bring it up."

BountyQuest's CEO, Charles Cella, says his company's discovery of prior art came too late to play a role in the InTouch suit. But BountyQuest's failure to affect the InTouch patent, combined with the market's apparent lack of interest in the bounty-hunting approach, may signal more serious problems with the whole idea of market-based reform for the thorny legal problems involving intellectual property. Despite tremendous media attention and Bezos' star power, BountyQuest's impact has been minimal. The U.S. Patent Office continues to rubber-stamp thousands of computer-related patent applications while BountyQuest largely remains relegated to the sidelines.

Cella is optimistic, convinced that his company will survive, and that it fulfills a useful and important role. But can it stem the tide of bad patents? Can BountyQuest, or any market-based approach, create the kind of change that so many have been clamoring for since 1999?

Other opponents of the current patent law status quo are proposing their own reforms -- and some of the new ideas also aim to solve the problem from below, by creating market-based alternatives that use the profit incentive to force positive change. In a way, they are holding on to one of the cherished dreams of the Internet bubble -- that in the New Economy, everything will be different, and more efficient, if we can just figure out the right online business model. But BountyQuest's track record over the last year also suggests that a different, more traditional approach may be necessary -- legislative reform to fix a broken system.

The story of BountyQuest begins with Amazon.com. In February 2000, the giant bookseller announced it had been granted a patent for its affiliate system. Bezos had already earned criticism for patenting the technique of one-click shopping, but the idea that he could also protect the concept of rewarding business associates who refer customers through a hyperlink struck many as absurd.

The Net reacted with electric ire. A week after Amazon won the affiliate patent, more than 1,300 people united at NoWebPatents.org, threatening to boycott Amazon unless it gave up its strategy of patenting "obvious" and widely used technologies. At the Web site run by book publisher O'Reilly & Associates, visitors posted about 3,500 responses to CEO Tim O'Reilly's critical open letter to Bezos. Amazon's own discussion boards brimmed with anti-patent messages.

At first, criticism focused on Amazon for what many felt was overreaching behavior. But over time, the patent system itself became the favored target. Taking their lead from documents like Pamela Samuelson's 1994 Columbia Law Review article "A Manifesto Concerning the Legal Protection of Computer Programs" -- which argued that software patents threatened to hinder innovation -- programmers and free software leaders began to turn the fight against bad patents into a crusade. Software patents "obstruct software development, and prohibit free software," explains Richard Stallman, founder of the Free Software Foundation.

"The only software patents that are not bad are the ones that will never be enforced," he adds.

Even those who took a more moderate line agreed that patent applications should only be issued after intense scrutiny -- particularly in regard to exploring whether previous solutions, or prior art, already existed. According to critics who included Greg Aharonian, publisher of the Internet Patent News, the U.S. Patent Office didn't take its job seriously enough. Computer-related patent applications spiked to 15,606 in 1998, up from 9,250 in 1997, reported Aharonian. But rather than increase the number of examiners to handle the load, the office seemed content to simply do less research.

"On average, issued software patents are missing three or four prior art examples that should have been cited," Aharonian said in a March 2000 interview. "Examiners don't know how to search effectively."

The Patent Office, then and now, denies Aharonian's claims. "There is no way that anyone can determine what art an examiner considered without looking at the file wrapper," says Brigid Quinn, a PTO spokesperson, referring to the complete body of documentation that accompanies each patent award. "Greg did not do that."

In the Patent Office's defense, Amazon's affiliate patent, and InTouch's two downloading patents -- cite at least a dozen sources, many of them available online.

But in an odd twist, by March 2000 Jeff Bezos had added his name to the list of agitators for change. In a letter posted on Amazon.com, he slammed the patent system and suggested several changes. Specifically, he sought the creation of a special set of laws to govern software and business-method patents, to shorten the life span of such patents and to create a database of prior art to help educate the Patent Office about existing innovations.

"Bottom line: fewer patents, of higher average quality, with shorter lifetimes," he concluded. "Fewer, better, shorter. A short name might be 'fast patents.'"

Enter Charles Cella, a practicing patent attorney. Cella didn't believe that the entire system needed fixing, nor was he necessarily opposed to the idea of software patents. He simply figured that there was a market opportunity in helping patent examiners determine what innovations really were novel and unobvious enough to deserve protection.

"There are a lot of bad patents out there, and there are a lot of valid patents out there," he told Salon in October 2000. "What needs fixing is the system of prior art."

Cella obtained investment from both Jeff Bezos and Tim O'Reilly, thus guaranteeing media attention and street credibility. And at first, all went well. Patent lawyers said they were willing to give BountyQuest a try. Free software geeks and civil libertarians welcomed the idea. O'Reilly even gave the site a high-profile kick-start, by agreeing to pay $10,000 to anyone who submitted prior art that invalidated Bezos' one-click shopping patent.

Amazon's involvement in BountyQuest proved to be a double-edged sword. First, the award-winning prior art submission invalidating InTouch's downloading patents raised eyebrows. Then, when BountyQuest announced that no prior art had been discovered undermining the one-click patent, others began to wonder.

Today Cella and O'Reilly dismiss the supposed scandal as an aberration and a distraction. "This is completely ridiculous," says O'Reilly, addressing the one-click patent issue. "Amazon had nothing to do with the judging on the submissions. BountyQuest sent me two big binders with everything that had been submitted. They told me the things that seemed most relevant to them, but I made the final decisions -- and in fact, I thought that a couple of things were more compelling than they did, as patent lawyers, so I awarded a prize even though they thought that nothing definitive had turned up."

But scandal or not, BountyQuest had larger problems. There just didn't appear to be a market for its service. Cella had hoped that BountyQuest's initial successes at discovering prior art would knock out at least one or two bad patents, which would bring the site more attention, not just from the press, but also from lawyers and the U.S. Patent Office. But the Hollywood climax or "splash cases," as Cella calls them, never came.

BountyQuest tried to overcome the inability to build momentum by cold-calling patent lawyers and trying to sell them on the idea of running a contest for one of their cases. But few have proved willing to bite.

Randy Lipsitz -- a partner at Kramer Levin Naftalis & Frankel in New York -- who remembers getting calls from BountyQuest, says he didn't use the company because he never had a client that felt satisfied with the old way of doing things. "There are experts out there who are knowledgeable in the field and you typically go to them," he says. "Prior art is information that needs to be mined. It needs to be put together like a puzzle. It requires expertise."

It also requires, suggests one BountyQuest competitor, a better appreciation of the whole market for prior art -- separate from the ideological goal of busting bad patents. Examples of prior art are commodities in and of themselves. If you can prove you've got prior art, you can sell it to a company wishing to bolster its own patent defense plans.

"BountyQuest was always a joke to those who understand prior art searching," says Greg Aharonian, who makes his living busting patents. "Professional searchers like myself typically bust patents for much less than what BountyQuest charges, and those patents we can't bust, well anyone smart enough to figure out how to find the prior art will know enough about the industry to go straight to the players -- law firms, companies -- and sell the prior art directly, cutting out the BountyQuest middlemen."

"Bezos and O'Reilly were never seriously interested in patent quality," he adds. "Bezos just used O'Reilly to help Amazon public relations. That Amazon ended up licensing the InTouch patent just shows how stupid the whole thing is."

Instead of relying on the patent market to adopt new ideas like BountyQuest's, government must intervene and create change, Aharonian says.

"One reform is to require applicants to do prior art searches, as opposed to the current rule that they disclose what they know," he says. "The current rule allows them to wimp out by saying, 'We didn't know because we didn't search.'"

Others are calling for more radical government reform. Richard Stallman, founder of the Free Software Foundation, argues Congress should exclude software from the patent system. "That," he says, "would really solve the problem."

Lawrence Lessig, Stanford law professor and author of several books on code and intellectual property, takes a slightly more moderate stance. He argues that the PTO should never have started approving cyberspace patents in the first place. By allowing patents to cover innovations in software, the PTO dramatically expanded the scope of intellectual property law largely through a form of institutional inertia. "Rather than reason, what governs the current patent debate is bias -- bias in favor of a system that seems right just because it seems old," he writes in his latest book, "The Future of Ideas." "But the relevant system is not old -- it is being expanded in ways that would shock lawyers of a generation ago."

To stem the tide of cyberspace patents -- which essentially give creators a government-backed monopoly on the idea that's patented -- the PTO should not necessarily just kill the possibility of software patent, as Stallman suggests. But before approving more applications, the agency should first have to prove that software patents serve the larger goal of encouraging innovation. "We will never know what benefit this regulation provides," Lessig argues, "until we begin to demand that the regulation prove itself."

And even after this test is held, assuming software patents pass, Congress should go further, Lessig argues, by expanding the laws of disclosure. Software inventors now only have to reveal an idea's general description; they should also have to make the source code public.

Few of these experts have given up on market-based reform. Lessig and other academic researchers, in fact, will soon launch Creative Commons -- a market-based alternative to the intellectual property system that will offer free, customized licenses for artists, writers, programmers and others who want to, say, allow students to download their program's source code but not large companies.

Eben Moglen, a Columbia law professor and counsel to the Free Software Foundation, also has an idea of his own. He suggests collecting enough software-related prior art in one place to gain the kind of momentum that BountyQuest lacks.

"My impression is that the [BountyQuest] model ('We'll ask the questions and wait for others to come up with answers') is less effective than ('Here's an immense database of answers waiting for questions, and by the way if you've got answers we don't have, contribute those too')," he says. "A running nucleus, as the successful project leaders will tell you, is the seed that gets development growing."

But will any of these ideas be enough to transform the system, to staunch the flow of bad patents?

Cella and O'Reilly still believe that market-based reform is the best option. Joshua Kaplan, InTouch's CEO and a regular defender of the patent system, also figures that companies like BountyQuest are far from useless, and possibly helpful.

But the government is already beginning to awaken from its slumber. The Federal Trade Commission held the first hearings on the patent issue Feb. 6. And while slow, some say that the kind of reform that government can bring might be just what the patent system needs.

Some problems -- such as the call for source code disclosure -- can only be accomplished through legislative action. And BountyQuest's experience reveals just how hard it could be for any market-based idea to gain traction.

"The intellectual property system is resilient, and hard to reform," says Mark Lemley, an intellectual property expert and law professor at Berkeley.

"BountyQuest was a great idea and may still prove to be, but you must keep it in context," says Lessig. "It is a great idea for an awful system."

Even Cella has tempered his rhetoric. While he says that he's still "extremely skeptical about legal or political reforms" because "they take forever, and they are almost always subject to the law of unintended consequences," he also no longer focuses so heavily on what BountyQuest can do for the system. Instead, he speaks of broader changes. "We alone aren't going to provide enough of a reform," he says. "There needs to be more."

Will those who followed Cella into the wild-blue yonder of market-based reform also start to take a softer line on civic activism? To truly fix the patent system, they'll have to do more than tout the Net's ability to transform the world. They'll have to do more than trust BountyQuest.

As Stallman puts it: "BountyQuest can do more good than harm -- provided we don't mistake it for a real solution."

By Damien Cave

Damien Cave is an associate editor at Rolling Stone and a contributing writer at Salon.

MORE FROM Damien Cave

Related Topics ------------------------------------------

Amazon.com Intellectual Property