Sometimes, the devil’s in the dicta.
The Supreme Court’s 6-2 decision in Commil v. Cisco rejects good-faith belief in a patent’s invalidity as a defense to induced infringement. It also features a surprising exchange between Justice Kennedy, writing for the majority, and Justice Scalia, dissenting (joined by Chief Justice Roberts), regarding the problematic proliferation of patent-assertion entities (“PAE”). This important issue of social and economic policy emerges as an afterthought forming Part III of the majority’s opinion, tangential to the analysis undergirding the Court’s decision. The majority takes pains to criticize PAE’s and recommend legal sanctions for frivolous patent litigation.
The Court didn’t have to include a section reminding district judges of their responsibilities. They state outright that the case raises no frivolity issues. Evidently, the Court wants to deter patent trolls and wants us all to know it. Yet, as the dissent observes, the decision facilitates future trolling.
What is a Patent Troll?
Remember the Three Billy Goats Gruff? The Norse folktale features a big, bullying brute on a bridge, threatening to devour the goats who try to cross. Only when the third and largest goat trounces him do the protagonists get fat on the green grass across the river.
Since the 1990s, this image has been borrowed to describe “patent assertion entities,” a type of “non-practicing entity.” These are firms that acquire and aggressively litigate (or threaten to litigate) technology patents, rather than practicing them. The moniker suggests that, like bridge-haunting trolls, PAE’s stand between society and its greener pastures. Many experts have concluded as much.
If the dissent is right, and the Court has just “increased the in terrorem power of patent trolls,” (the first use of this pejorative in Supreme Court history), there will be negative consequences for scientific innovation – a primary goal of patent law.
The Case
Legal Background
There are three roads to liability for infringing upon a patent, laid out in 35 U.S.C. § 271(a)-(c): direct infringement, induced infringement, and contributory infringement. Direct infringement is a strict liability offense, disregarding an infringer’s knowledge or intentions, if they make, sell, or use patented material without permission from the patent holder. Induced infringement, however, requires showing that the defendant “actively induced” infringement by another party. Under Supreme Court precedent, the inducer must know of the patent, and intentionally encourage actions that they know to infringe thereon.
Because of these knowledge requirements, one defense to induced infringement liability is good-faith belief that the induced actions aren’t infringement. The Supreme Court granted certiorari to rule on whether good-faith belief in the patent’s invalidity would similarly suffice.
Factual Background
The opinion doesn’t focus on the technical details of the patent, a type of wireless networking innovation. Nor does it dwell on the fact that Commil USA, the ruling’s immediate beneficiary, is a non-practicing entity. But the latter point provides some context for the decision’s seemingly unprompted discussion of patent trolls in Part III, and Justice Scalia’s blunt rejoinder.
Commil was originally an Israeli startup founded in 2000. In financial ruin, the company fired its employees and sold its patents “for a few hundred thousand dollars” in 2005. Thus was born Commil USA, the patent-holding company that sued Cisco in 2007. Litigating or threatening litigation on its acquired patents is Commil USA’s raison d’être, and in the Eastern District of Texas they won a $63.7 million verdict.
On appeal in the Federal Circuit, Cisco challenged, inter alia, the district court’s suppression of evidence as to their good-faith belief in the patent’s invalidity. A divided court held “it is axiomatic that one cannot infringe an invalid patent,” and thus it is impossible to knowingly infringe what one doubts to be valid. In partial dissent, Judge Newman opined that “whether there is infringement in fact does not depend on the belief of the accused infringer that it might succeed in invalidating the patent.”
Holding
In Commil, the Court agrees with Judge Newman. While rejecting Commil USA’s aggressive reading of §271(b), which would impose liability for anyone who induced infringement on a patent they knew about – even without knowledge that the induced acts constitute infringement! – the majority also rejects the Federal Circuit’s “axiom.” The Court refuses to conflate the distinct defenses of patent invalidity and noninfringement, which are listed separately in the Patent Act, and only accepts “good-faith belief” in the latter.
One problem with the proposed defense is that patent law rests on a presumption that patents are valid. Plaintiffs needn’t demonstrate their patents’ validity to bring suit – that’s for infringers to challenge by “clear and convincing evidence.” But if a reasonable belief in invalidity were a defense to certain actions, this presumption would be undermined, constantly requiring plaintiffs to prove that defendants “couldn’t reasonably believe” their claims. This would allocate evidentiary burdens unfairly, and prevent timely and orderly administration of law by importing validity questions into the (separate) infringement inquiry.
Likewise, allowing this defense would encourage every infringer to concoct arguments about patent invalidity: to escape liability they wouldn’t need to demonstrate invalidity, just sufficiently justify prior belief in invalidity. This would greatly extend litigation.
Finally, the Court notes that numerous legal processes already allow firms to challenge patents’ validity before risking infringement. Firms should employ them, or risk the consequences.
Part III
Rather abruptly, the Court digresses to discuss patent trolls.
The Court is well aware that an “industry has developed in which firms use patents not as a basis for producing and selling goods but, instead, primarily for obtaining licensing fees. Some companies may use patents as a sword to go after defendants for money, even when their claims are frivolous. This tactic is often pursued through demand letters, which “may be sent very broadly and without prior investigation, may assert vague claims of infringement, and may be designed to obtain payments that are based more on the costs of defending litigation than on the merit of the patent claims.” This behavior can impose a “harmful tax on innovation.”
No issue of frivolity has been raised by the parties in this case, nor does it arise on the facts presented to this Court… (internal citations omitted)
The Court proceeds to remind us that district courts have the “authority and responsibility” to impose sanctions for frivolous litigation. It concludes that these existing safeguards, along with the other avenues available to those who doubt patents’ validity, justify its holding.
Dissent
Justice Scalia’s dissent is brief but sharp. Decrying the Court’s use of social consequences to determine whether or not to “create” a “new” defense to induced infringement, he would have the Court limit its analysis to interpreting the statute. In turn, that analysis would accept the syllogism that one cannot knowingly infringe on what one doubts to be a valid patent.
In his last comment, however, he predicts that the Court’s opinion will sharpen the swords of the patent trolls discussed in Part III. Not only is the majority wrong to base their holding on projected social consequences, their holding magnifies an acknowledged harm.
Potential Consequences
Justice Scalia’s prediction has merit. As noted by the Electronic Frontier Foundation (who submitted an amicus brief supporting Cisco’s position),
Many of the worst patents out there are easily wielded as weapons precisely because they are very broad (meaning they are often infringed) but also likely invalid. . . . [P]atent trolls with invalid patents can still use this decision to extort money since they might have a stronger infringement claim and can strategically dismiss lawsuits to evade having their patent invalidated by a court when actually challenged.
Furthermore, as their brief argued, this will encourage patent applicants to use unclear descriptions of inventions to claim broader monopolies. After all, if others can’t tell by reading the patent whether it’s valid, that’s now their problem! This will force the Patent Office to spend more time and resources reviewing applications with heightened care.
Conclusion
If Part III (and its terse rejoinder in the dissent) were excised from Commil v. Cisco, patent law would allow and disallow the same defenses to induced patent infringement. The Court, mindful that shifts in the balance of power toward patent plaintiffs don’t only help the “good guys,” expended precious judicial ink to signal its desire to slay the trolls.
Yet, refusing to interpret the Patent Act in a logico-semantic vacuum, the majority had to weigh judicial administrability and plaintiffs’ ability to defend their intellectual property on the one hand, against an acknowledged social ill. So they accepted a bifurcated definition of the scienter requirements, one which still enables powerful “good-faith belief” defenses in the relatively clear arena of infringement doctrine, but rejects it in the notoriously nebulous realm of validity doctrine. And they reiterated along the way that invalid patents and illegitimate plaintiffs should be challenged vigorously.
According to the dissent, this isn’t their job, and they’re doing it badly. Just read the statute and accept its inexorable logic. If the Patent Act is thereby rendered unwise, there’s another branch of government who are supposed to do something about that.
But maybe in demanding that lower courts and potential infringers take a more active part in discouraging baseless IP litigation, the majority partially agrees: America can’t look to them alone to chase the trolls off our bridge. After all, they are the smallest of our three federal branches gruff.
Alex Schindler is a J.D. candidate, 2017, at NYU School of Law.
https://youtube.com/watch?v=K9WmxwgW2J0
