In June 2012, the New England Federal Credit Union in Williston, Vermont, received a vaguely worded letter from a company called Automated Transactions, warning the firm that it might be using teller-machine technology that infringed on Automated’s patents. The letter had several mistakes, New England Federal’s CEO John Dwyer informed a U.S. Senate committee last December, starting with its addressee: Dwyer’s predecessor, who had departed two years earlier. Automated and its lawyers, Dwyer complained, had “performed little or absolutely no research prior to demanding money from our credit union”—in effect, it had sent a “form letter.” Still, Automated proposed “a special one-time limited offer” to license automated teller-machine technology to New England Federal for $2,000 per ATM. The patent firm later hiked the fees it wanted, and then sued New England Federal, seeking triple damages. “The case has been a costly and distracting headache,” Dwyer continued, “from a company that has made a business out of what, in my opinion, in another context, might look like extortion.” All this, even though a judge had already invalidated several Automated patents, as the credit union’s research would uncover.
Automated is what many now call a “patent troll”—a firm that holds patents of dubious merit, with which it produces no products but that it instead uses to demand licensing fees from businesses. Patent trolls have become so controversial that Vermont senator Patrick Leahy, chair of the Senate Judiciary Committee that held the December hearing, observed, “You don’t have to be a patent expert to know what is going on. I mean, this is as close to robbery as you can think of.” Leahy promised action on a bill to end some of the abuses that Dwyer described, but barely five months later, the senator deep-sixed his own legislation, despite bipartisan support for it. Reportedly, Senate Majority Leader Harry Reid rejected the Leahy bill because of intense opposition from the powerful trial lawyers’ lobby.
In fact, the opposition ran deeper than the trial bar, threatening future patent reform. Flaws in our patent system, which the distinguished appellate judge and law professor Richard Posner dubs “dysfunctional,” have transformed the technology market, making ceaseless litigation lucrative not only for Automated and patent trolls like it, but for others, too. The winners include universities, which increasingly sell otherwise unmarketable patents to shake-down trolls; established companies, which launch patent suits to hamstring competitors; and, of course, the lawyers. The losers: innovation and the American economy.
On August 17, 1787, James Madison, rising before the delegates meeting in Philadelphia to draft a constitution for the new United States, suggested that the document should contain some provision to protect intellectual property. In Federalist 43, Madison noted that Great Britain granted such rights, which helped to advance the public good by encouraging innovation. Madison’s efforts culminated with Article I, Section 8 of the U.S. Constitution, authorizing Congress to promote “the progress of science and the useful arts,” by granting exclusive rights to authors and inventors. The Patent Act of 1790 represented the first federal law codifying the assigning of these rights. Under it, patents became “a just reward to ingenious men, and highly beneficial to the public,” in the words of nineteenth-century Supreme Court Justice Joseph Story.
Nevertheless, early controversies foreshadowed patent trolling. In the late nineteenth century, the invention of the sewing machine triggered a notable patent battle. Various inventors—including Elias Howe, Charles Morey, and Isaac Merritt Singer—came up with different versions of sewing machines, seeking a commercially viable product. Howe’s was a marketplace dud, but he proceeded to demand licensing rights from the other machine makers—and then launched several lawsuits, which prompted his competitors to countersue. Eventually, the squabbling manufacturers resolved to band their patents together into a single pool and grant licenses to one another. Howe never manufactured sewing machines for profit but wound up collecting some $2 million in royalties from sales of the machines by others.
Our twenty-first-century patent wars are a supersize version of the sewing-machine conflict. They have their origins in changes to patent law that have occurred over the last several decades, especially concerning computer software. The U.S. Patent and Trademark Office long resisted granting patents on software. As one federal court described the Patent Office’s traditional position, “a programmed general-purpose computer could never be viewed as patentable subject matter”—the programming itself was simply too general and abstract. Patent law clearly stated that an application for a patent had to be concrete in its description of any innovation and “shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art . . . to make and use” the invention.
But as software became more common, courts reversed course. In 1981, the Supreme Court ruled that a software program that controlled a physical process—in the case in question, a computer-operated precision-molding machine—could be part of a patent. In 1994, a federal district court went further, overruling the Patent Office to allow a patent on software that ran a computer-driven digital display. By the late 1990s, a flood of patents covered computer programs. Today, the Patent Office approves a staggering 40,000 software programs a year.
Many of these new patents explain computer processes to solve a problem or reach a desired end, without defining any actual device or unique code that makes it happen. It’s the equivalent, says Stanford law professor Mark Lemley, of a pharmaceutical company getting a patent for “an arrangement of atoms that cures cancer.” Even many software designers recoil at such patents. In March 2012, Andy Baio, cofounder of Upcoming.org, a now-defunct events calendar that Yahoo acquired in the mid-2000s, described in Wired how the tech giant had asked him for information on anything potentially patentable that he’d created for the site. “I’ve always hated the idea of software patents,” wrote Baio. “But Yahoo assured us that their patent portfolio was a precautionary measure, to defend against patent trolls and others who might try to attack Yahoo with their own holdings.” Yahoo then filed four patents based on Baio’s work, which he describes as “nonsensical documents that I still can’t make sense of.” One patent, he observed, merely asserted a claim for “[t]echniques . . . for notifying a member of a social network system of those changes, within the social network system, that are of most interest to the member.” The firm could use the patent, Baio worried, to demand licensing fees from every social-media website with a news feed. Baio became especially concerned when a struggling Yahoo began “weaponizing” patents to sue others, including Facebook. “Unable to innovate, Yahoo is falling back to the last resort of a desperate, dying company: litigation as a business model,” he noted. Worse, Yahoo could one day auction off its patent portfolio to still more aggressive trolls.
Many cases winding through the legal system involve similarly abstract patents. The Gooseberry Natural Resources company got a patent for creating and distributing news releases through e-mail—a process that required no true innovation. It then sued several big players, including Yahoo, MSN, and Reddit, for infringement. Some defendants settled out of court by agreeing to pay a licensing fee, viewing the cost of trying to invalidate the patent as prohibitive. Another defendant, Fark.com, went ahead with the case and easily defeated the troll’s claims, but its legal bill was massive. “One of the major problems with patent law is that, in the case that you are sued by a patent troll, the burden of proof that you did not infringe on the patent is actually on the defendant,” observed Drew Curtis, Fark.com’s founder.
Another case concerns Personal Audio, a company that dabbled in dispensing audio content on tape during the 1990s, when it patented the idea. Though Personal Audio failed to make money with its original concept, which was to distribute tapes of magazine stories read aloud, it now claims to own the rights to any “system for disseminating media content representing episodes in a serialized sequence.” That includes, among other things, the proliferating practice of Internet podcasting. Personal Audio, which makes no products today, has accordingly demanded licensing fees from the biggest podcasters, including $3 million from comedian and commentator Adam Carolla. The claim might seem outlandish, but the troll had already scored against a bigger target: Apple. In early 2013, a Texas jury determined that the music players in the iPod, iPad, and iPhone infringed on another Personal Audio patent, this one for an audio player that lets listeners create their own playlists; Apple had to fork over $8 million. Carolla vowed to fight his suit and asked listeners to chip in for his defense, but the comedian’s legal bills exceeded the nearly half-million dollars he raised. (See sidebar.) Even though Personal Audio recently agreed to drop the suit, after discovering that Carolla didn’t make big bucks from his podcasts, the troll doesn’t owe him a dime.
Earlier this year, an apparent patent troll called Red Pine Point filed suit against Magnolia Films, charging that the company was violating a patent that it held on the concept of selling movies on demand, prior to their theatrical release, to Internet viewers. Magnolia, partially owned by entrepreneur Mark Cuban, began marketing its movies that way in 2004. In 2006, Red Pine Point obtained a patent for the practice, which listed a lawyer and Red Pine executive, Philip Lyren, as coinventor. The vague patent claims purview over “methods, apparatus and systems for marketing and selling feature length movies over one or more networks . . . and includes placing, at plural different Internet locations, advertisements to purchase a feature length movie (FLM) before the FLM is available to buy.” The tech publication Ars Technica sarcastically dubbed Lyren “inventor of the week.” Cuban, who has funded the “Chair to Eliminate Stupid Lawsuits” at the Electronic Frontier Foundation, a digital advocacy group, wrote on his blog: “Could there be a more ridiculous patent ever issued?”
MIT economist Catherine Tucker has demonstrated how trolls, exploiting such vaguely worded patents, can snuff out innovation. In one study, Tucker looked at medical-imaging technologies, which hospitals have adopted to store and access electronically the records from CAT scans, ultrasounds, and other tests that previously existed only on paper. In 2005, a patent troll known as Acacia Research purchased a patent from the 1990s for a “medical image storage, retrieval and transmission system.” Though no one was manufacturing devices under the patent, Acacia claimed that any process that now accomplished the same digital-imaging transmission infringed on its intellectual property. It warned hospitals and makers of medical-imaging systems to expect lawsuits. Sales in the field plunged by half and research funding fell, as imaging firms stopped releasing new products or updating existing ones. “Vendors did not want to run the risk of being found guilty of ‘wilful infringement’ in the patent suit and being held liable for treble damages,” observes Tucker.
The trolling has produced a free-for-all in the courts. Patent litigation spiked 85 percent from 2006 through 2012—a rise entirely attributable to suits filed by so-called nonperforming entities, another designation for trolls (they’re “nonperforming” because they make no products). A 2012 study by Boston University law professors Mike Meurer and Jim Bessen estimated the annual cost of patent litigation in the United States to be $29 billion. The broader effect on the economy, including the diversion of resources away from productive investments and innovation, may reach $80 billion yearly, they say.
Patent plaintiffs often shop for sympathetic legal venues. Many trolls, for example, have opened offices within the jurisdiction of the federal Eastern District Court of Texas, located in tiny Marshall, Texas, where courts have proved unusually friendly to patent holders and judges are less likely to dismiss seemingly frivolous suits. Of cases that go to trial, Marshall juries decide in favor of patent plaintiffs 79 percent of the time, compared with 59 percent nationwide. Of the 6,000 patent lawsuits filed in the U.S. last year, one in four landed in Marshall, according to the Electronic Frontier Foundation. Marshall’s small-town atmosphere—the town’s population is only 24,000, with a median household income of just $33,000 annually—may help explain why juries there are so partial to plaintiff claims, especially when, as is often the case, the defendant is a big company accused of stealing an invention.
Small businesses are particularly vulnerable to all this litigation, since the legal costs of defending themselves are potentially ruinous. A 2012 study by Santa Clara University law professor Colleen Chien estimated that 55 percent of troll-sued firms have less than $10 million in revenues, while the average cost of litigating against a patent claim was $837,000. Trolls know that small firms don’t want to go to court, which helps explain the vast increase in demand letters, like the one that New England Federal received from Automated Transactions, offering to settle an infringement case quickly for a more “reasonable” sum.
Given the price of battling patent trolls in court, it’s not surprising that few companies try. In a 2013 study, George Mason University economist Shawn Miller estimated that more than half of software patents would wind up invalidated if legally challenged. But just challenging a patent through the Patent Office would likely cost between $100,000 and $400,000, and much more if the case goes to court. Courtroom failure can mean that the defendant exposes itself to hefty damages, too.
Start-ups are a primary target of the shakedowns: 35 percent of the new technology firms that Chien surveyed had gotten a demand letter. Even settling out of court with the trolls often turned out to be expensive—the average payout was $340,000, Chien estimates. For new firms seeking financing, a demand letter can thus be a death knell. “No one wants to invest in a company where founder time and investor money is going to be bled to patent trolls,” Chien says.
Many trolls don’t bother going after the producers of technology products but instead threaten small users of those products—companies that have no stake in racking up scary bills to defend someone else’s intellectual property. Automated sought licensing fees from dozens of small financial institutions like New England Federal and other retail outlets that use, but don’t make, ATMs. Similarly, the troll MPHJ fired off more than 16,000 letters to firms claiming that it owns the patent for scanners—the common devices that transform paper documents into digital files—and saying that it had to be paid up to $1,200 per employee as its licensing fee. A giant troll, Innovatio IP Ventures, contends that it holds the patent on wireless Internet technology. It has sent off an estimated 13,000 letters in recent years to small users, including coffee shops, hotels, and retail-food chains like Panera Bread, demanding $2,300 to $5,000 in licensing fees per outlet using Wi-Fi.
Trolling has proved so lucrative that litigation-minded firms have started to accumulate patent portfolios. This has set off bidding wars for patents, some of them thought commercially worthless before their purchase. The most successful trolls have assembled hundreds—sometimes thousands—of these previously unused patents, at deeply discounted prices. One of the largest portfolios belongs to Intellectual Ventures, which has purchased or licensed the rights to between 30,000 and 60,000 patents, according to a 2012 Stanford Law Review article by Tom Ewing and Robin Feldman. Intellectual Ventures has even engaged in what Ewing and Feldman call a new form of patent “privateering”: licensing its patents to other trolls, who then sue companies, sharing any “profits” with the giant firm.
Looking to bring in funds, technology companies—the frequent victims of trolls—have gotten into the shortsighted practice of selling patents to portfolio assemblers. In 2009, for instance, chipmaker Micron Technology sold a portfolio of 4,200 patents in chip technology, photo imaging, telecommunications, and search engines to Round Rock Research, a company controlled by a patent lawyer; Micron justified the sale as an attempt to recoup research costs.
The amassing of portfolios by trolls has generated a windfall for universities. The schools became leading patent players with the 1980 passage of the Bayh-Dole Act, allowing them to secure patents from any research they undertook with federal money (previously, the federal government grabbed those patents). Universities obtained about 250 patents yearly, pre-Bayh-Dole; now, they average about 5,000 annually. Bayh-Dole has won praise for spurring innovation by quickly getting important research into the hands of businesses. But more recently, universities have resorted to controversial sales and auctions, seeking big dollars from patents that aren’t commercially viable. “Time and time again when I talk to people in a variety of industries, their view is that universities are the new patent trolls,” Stanford Law’s Mark Lemley observes. One patent lawyer, speaking at a Fordham Law conference, called universities “crack addicts,” hooked on patent royalties. Ewing and Feldman’s Stanford Law article identified 50 schools—including CalTech, Clemson, Duke, and the University of Texas—as having sold or licensed patents to Intellectual Ventures. An executive of Ocean Tomo, a patent-auction firm, told the journal Nature last year that auctions featuring university patents make up 20 percent of the firm’s business.
The money that universities can generate in this way can be massive. In one illustrative case beginning in the 1990s, Michael Doyle, a researcher at the University of California at San Francisco, claimed to have invented websites’ ability to use interactive features. Doyle patented his idea, launched a company called Eolas, and agreed to share licensing fees with the university. Eolas then began suing major Internet players, including Microsoft in 1999. Though several Internet pioneers, including World Wide Web creator Tim Berners-Lee, testified that others had come up with these features well before Doyle, a jury in 2006 awarded Eolas $521 million against Microsoft. The University of California later disclosed that it obtained some $30 million as its share of a subsequent settlement. Six years later, another jury invalidated the Eolas patent in a case brought against it by Google, Amazon, and Yahoo—but not before Eolas and the university had won judgments against or licensing deals with nearly two dozen other companies that had declined to fight in court.
Well before Senator Leahy killed his patent-reform bill, universities joined with trolls and trial lawyers in fighting key provisions of the legislation. Universities were particularly opposed to a “loser-pays” clause, which would have let judges in patent cases reward defendants with legal fees paid by the plaintiffs. Similar provisions are common features in legal systems around the world, but universities worried that instituting loser-pays would undermine their right to sue for patent infringement. That claim helped give political cover to Senator Reid and the trial lawyers—Reid’s top political donors, who have an obvious interest in keeping the litigation merry-go-round spinning.
You Didn’t Invent That
Patent number 8,112,504 is filled with the semi-technical yet ultimately hollow language typical of many software patents. The document declaims exclusive rights to “A media player for acquiring and reproducing media program files which represent episodes in a series of episodes as said episodes become available.” The diagrams illustrating the so-called invention consist of rudimentary drawings—including one merely containing six boxes linked together by arrows and bearing vague instructions, such as “Establish Account.” Yet from this elementary explanation of a system, the patent holder, Personal Audio, has declared rights to the entire field of podcasting and other content distributed online and has demanded licensing fees from various media companies.
Patent 8,112,504 may well be the most infamous patent in America today, after one of Personal Audio’s targets, comedian Adam Carolla, fought back, using his daily podcast to campaign against the firm and against patent trolls in general. Urging podcasters and listeners to band together to save the industry, Carolla began a campaign to self-finance his legal fight against the troll, raising $478,000 from 12,400 mostly small contributions on Fundanything.com.
In July 2014, some 18 months after it had filed suit, Personal Audio offered to drop the case against Carolla, after learning that the comedian didn’t make enough money from podcasting to justify the $3 million fee it had demanded. But a defendant should not have to raise half a million dollars for legal fees before a plaintiff realizes that he isn’t worth suing. Our patent court system and the judges who supervise it let defendants get caught in this expensive process, known as discovery, in which a plaintiff can demand voluminous documents, at little cost to itself, from a defendant, which accumulates big legal bills trying to comply before a case even comes to trial. Patent trolls depend on the burdensome process to scare firms into paying them licensing fees instead—because, as Carolla’s case shows, even victory can become massively expensive.
Though Carolla initially declined to settle with Personal Audio, hoping to win in court, a month later he opted out of the litigation. Going to court would have cost him several hundred thousand dollars more in legal fees. On average, he generated just $38 per contributor on Fundanything.com. Absent some big new donor, the comedian would thus have needed between 8,000 and 10,000 additional contributions, or nearly as many as he’d already collected, to see the case through—and he may have lost.
In settling, Carolla waived two important options. Had he prevailed in court, he could have asked the judge to compensate him for his legal fees, on grounds that Personal Audio’s suit was frivolous. That’s a long shot, given the small number of cases where judges award such fees to defendants, but if a judge holds a troll liable, it serves as a deterrent to future litigation. Instead, Personal Audio will bear no responsibility for the burden it needlessly imposed on Carolla. Second, at trial, Carolla could have also asked a jury to invalidate the Personal Audio patent. That would have freed podcasting and other forms of Internet-distributed media from Personal Audio’s claims. With its patent intact, Personal Audio is still suing a handful of TV networks, alleging that their distribution of video content episodically over the Internet also violates the firm’s “invention.”
The Electronic Frontier Foundation is challenging Personal Audio’s patent through the U.S. Patent Office’s review process. Trying to intimidate EFF in March, Personal Audio sent the nonprofit a subpoena demanding numerous documents, including details on its donors, but a judge quashed it. A hearing on the validity of the patent itself isn’t scheduled until April 2015. The clogged case docket is another factor favoring the trolls: our entire patent system, processing some 40,000 new and often questionable software patents every year, lags massively behind in adjudicating disputes, and cases take years to resolve. Patent trolls count on it.
Big companies that innovate and make commercially successful products have seen how industry players profit from the dysfunction of today’s technology-patent markets. Many of these firms, especially in software, have faced dubious patent lawsuits themselves. Their response, unfortunately, has often been like Yahoo’s: to join the game as aggressors. Some companies have poured billions of dollars into accumulating their own patent portfolios and litigating to stop competitors from joining their fields.
Apple may be the most striking example. The Cupertino giant was an early patent-troll target because its most popular products—the iPod, iPhone, and iPad—aren’t a single invention but instead multiple innovations packaged together, some conjured in-house, others designed by outsiders. Many activities that users carry out with Apple’s devices, such as listening to music or answering e-mails, were familiar. But by the time the iPad and iPhone hit the market, Jobs and Co. had created new ways of doing them. That wasn’t enough to protect Apple from trolls asserting that the firm owed them licensing fees based on older patents—some written so broadly that they seem to circumscribe virtually all future innovation. It was just such a patent from the 1990s—a vaguely described “audio program player including a dynamic program selection controller”—that enabled Personal Audio to sue Apple successfully. A Singapore firm, Creative Technology, held a similar broad patent for a “portable music playback device” that forced Apple to cough up $100 million in extra licensing fees for its music players.
Apple thus decided to protect its smartphone franchise by extensive patenting and ferocious patent litigation—igniting what some have dubbed the smartphone patent wars, which even Apple CEO Tim Cook concedes are “maddening” and “a waste.” Apple has launched multiple, massive suits against Samsung, Google-owned Motorola Mobility, and HTC, a Google partner in smartphones. In one immense Apple suit against Samsung, jurors considered hundreds of highly specific technical issues, including whether Samsung had infringed on an Apple patent that allows a smartphone to distinguish between a swipe across its face by a single finger or by two fingers. Samsung countersued Apple, demanding $422 million. A jury awarded Apple $1 billion and nothing to Samsung, but more lawsuits and appeals followed. Meanwhile, others have entered the fray. Motorola has sued Microsoft in several countries, claiming that the Windows maker owes it licensing fees for using Motorola-patented video technologies. Microsoft has countersued, charging Motorola with seeking unreasonably large fees for basic industry technology.
The litigation frenzy has angered some judges. In June 2012, Judge Posner refused to let a huge case, involving suits and countersuits by Apple and Motorola Mobility, to come to trial. Posner argued that neither side had demonstrated that it faced substantial harm from the other’s actions and thus deserved a “right to relief” and that the cost of such a trial outweighed any public good that might come from it. U.S. District Judge James Robart was similarly frustrated by Microsoft’s and Motorola Mobility’s patent face-off. “The conduct of both Motorola and Microsoft has been driven by an attempt to secure commercial advantage, and to an outsider looking in, it has been arbitrary, it has been arrogant and frankly it appears to be based on hubris,” he wrote in 2012. “The court is well aware,” Robart added, “that it is being played as a pawn in a global industry-wide business negotiation.” A Stanford University study estimated that between 2011 and 2013, smartphone companies spent $20 billion on patent rights and litigation. Apple and Google now spend more on patents than on research and development.
The road to a fairer and better patent system, absent congressional common sense, will require the cooperation of courts, patent officials, industry groups, companies, and the states. The first priority is to limit the scope of the demand letters issued by patent trolls. Here, state leadership is already evident. Nearly a dozen states have passed laws seeking to curtail these missives, and new statutes are pending in at least a dozen more. Legislation in Vermont and Oklahoma requires that firms sending demand letters be specific about the patents that they claim to own and about how the recipient is infringing on them. Legislation in other states establishes penalties for false patent claims—when, say, a troll demands a licensing fee for an invalidated patent. Some bills would let defendants recover damages and attorney fees after getting hit with misleading demand letters.
Federal courts could undertake some helpful reforms, as well. One part of the squelched Leahy bill sought to limit abuse of the discovery process, as when plaintiffs request thousands of documents, many of little relevance, before going to trial, at great cost to the company they’re suing—an incentive for firms to settle out of court. But even without a new patent law, federal judges can constrain outrageous and irrelevant discovery requests, or cut litigation short if it becomes clear that the defendant isn’t guilty of infringing on a valid patent. Judges can also push for early mediation between plaintiffs and defendants, before discovery expenses mount.
Judges already have the power, too, to award attorneys’ fees to defendants for frivolous patent claims. But too few judges have granted these awards. From 2005 through 2011, Chien calculates, judges imposed attorney fees in just 194 of the thousands of patent cases filed. But in April, the U.S. Supreme Court, in a unanimous decision, rebuked the Federal Circuit Court of Appeals, which handles most patent cases, for setting too high a bar for awarding attorney’s fees, and ordered it to expand the grounds on which a defendant could claim fees. The Supreme Court also limited the circuit court’s power to overturn fee awards made by lower courts.
The courts also can help narrow the scope of technology patents. The Supreme Court has recently instructed lower courts that abstract ideas aren’t patentable. And Judge Posner argues for a separate category of patent on digital products that would expire within five years, instead of the 20-year term of most patents, acknowledging the pace of change in high tech. Other reformers have called for outright abolition of software patents. Last year, New Zealand ended software patents, claiming that a computer program is “not an invention.”
Finally, industry groups and big companies could begin to restrain themselves. Berkeley Law scholars formulated a new legal arrangement, known as a defensive-patent license, that might serve as a model. Under the plan, inventors and firms agree to place their patents in a large pool, along with those of like-minded inventors. Everyone in the pool would enjoy the right to royalty-free use of the others’ patents—but license holders would agree not to use their patents to launch aggressive litigation. The patent rights would only be used defensively, to protect against infringement claims by others. The first of these licenses is set to launch this fall.
Federal legislation of the sort that Senator Reid short-circuited in May would go a long way toward ending some of the more prominent abuses in patent law. But such a solution may be impossible as long as Reid, with his trial-lawyer supporters, remains Senate majority leader. Other actors will have to work on several fronts to clean up the patent-litigation mess clogging our courts—and stifling our economy.