HBO's Silicon Valley: a Legal Critique - (Part Three)
In this final installment on our series about HBO’s Silicon Valley, we return one more time to the legal entanglements that challenge Richard Hendricks and the Pied Piper crew and ask how authentic the portrayal of these issues really is. In Part One, we discussed employee IP assignments and restrictive covenants. In Part Two, we looked at the Children’s Online Privacy Protection Act. In this article, we turn to two areas raised towards the end of the fourth season: patent trolls and copyright infringement for musical compositions.
Stuart Burke Patent Troll
Burke, it turns out, is not a software developer. He has no technical background whatsoever. He is (cross yourself, toss a pinch of salt over your shoulder, and spit three times) a lawyer. The patent at issue was simply purchased. LaFlamme explains that his colleague’s modus operandi is to scoop up promising patents from their inventors and then leverage this secondhand IP to threaten startups with frivolous lawsuits unless they license his “technology.” Burke selects targets by starting at the bottom of the list of bestselling apps and working his way up, strengthening his case and increasing his demands as he goes. Pied Piper is lucky to have been hit up first, LaFlamme counsels Richard. $20,000 is a bargain. Patent trolls are controversial, starting with the terminology. Plaintiffs smeared as patent trolls seldom have a legitimate product of their own. Thus, the more neutral terms, Non-Practicing Entity (NPE) or Patent Assertion Entity, are sometimes offered as less emotionally charged synonyms for troll, usually by those who think that NPEs get a bad rap. However, critics of patent trolls are often less concerned about the failure of NPEs to commercialize their own technology than about other telltale troll-like behaviors, such as the assertion of overbroad or low-quality patents and the reliance on the ruinous costs of patent litigation to extract unmerited settlements or “licensing fees.” Whatever the defining characteristic of a patent troll, though, Stuart Burke clearly fits the bill. Not only does he not make or sell anything, he is pressing a patent of questionable validity, leveraging his victims’ unwillingness to litigate and generally using the courts in a way that feels more like a business model than a legitimate vindication of his rights as a patent owner.
Against his attorney’s advice, Richard decides he wants to slay the troll. He attempts to rally other likely targets of Burke’s trolling to pool their resources, only to have his fellow entrepreneurs turn around and settle. Richard’s disloyal allies are rewarded for their cooperation with discounted licensing fees, while Pied Piper is punished with increased demands. Only by threatening to blow up Burke’s sideline exploiting similarities in musical compositions (more on that subplot below) is Richard able to defeat the troll, though even this victory turns Pyrrhic when Pied Piper receives LaFlamme’s $22,000 bill for helping set up Richard’s ill-fated anti-troll coalition. It would be unusual for an individual attorney to be the sole party in interest behind a Patent Assertion Entity, both because it raises ethical concerns and because a single lawyer may not have the resources to fund a patent infringement suit, the costs of which can run into the millions of dollars. But, otherwise, Silicon Valley’s depiction of the patent troll phenomenon is pretty realistic.
Patent Assertion Entities will usually start with a demand letter just like Pied Piper receives. The demand will often be presented as a “licensing fee” and might be for a lump sum, as Burke proposes, though (perhaps to appear more like a legitimate license rather than naked extortion) might instead be in the form of an ongoing royalty scheme. And LaFlamme is correct that $20,000 is relatively cheap. He estimates that Pied Piper will be out at least $100,000 to hire a patent litigator, which is, if anything, optimistic. It is also a standard strategy among patent defendants to organize a “joint defense group” to confront a common antagonist, just as Richard attempts. However, the practice of patent trolls “going up a list,” allowing identification of future targets, is, as far as I know, a narrative device. Joint defense groups are usually comprised of businesses that contemporaneously are sued or receive demand letters. Similarly, the patent troll tactic of obtaining small initial settlements to shore up their claims is grounded in fact. The idea as presented in the episode is that licensing agreements increase the value of the patent, thus justifying ever-growing demands as the troll moves “up the list.” This is a dubious assertion from a legal perspective, but it’s not one that the show invented. Patent demand letters will often trumpet past licensing deals as proof of the patent’s legitimacy and value (often calling out the licensees by name if the terms of their settlement agreements allow it), and it’s not unusual for IP attorneys to rationalize aggressive enforcement as a way for clients to protect the value of their IP. To be sure, there’s probably a psychological value in a patent troll being able to prove its willingness to litigate and point to other companies who have folded rather than fight. But the impact of past “victories” is often overstated. Settlements do not prove a patent’s validity, and licensing deals do not prove its value.
As far as Pied Piper’s liability goes, we probably don’t know enough about Burke’s patent to speculate. Richard is certain it’s unenforceable. LaFlamme agrees it’s “probably not” but seems fundamentally uninterested in the question.
Richard’s frustration, which accurately reflects the frustrations of many patent defendants, particularly in the information technology sector, is the patent’s breadth. A patent is comprised of several sections, including a detailed description of the invention (usually called the “specification”), but the only section that really matters for purposes of infringement is the set of numbered clauses at the very end called the patent “claims.” Richard quotes the patent as covering “storage of media files on a network,” though that sounds more like the patent’s title than one of its claims. Nevertheless, it’s true that, whatever details might be provided in the patent specification, patent claims (especially computer-implemented method claims) are often drafted in broad, technology-neutral terms that exasperate engineers. Patent claims are subject to extensive examination by the U.S. Patent and Trademark Office, so it’s generally a mistake to assume that the claims in an issued patent are invalid or unenforceable across-the-board. However, the Supreme Court’s 2014 decision in Alice Corp. v. CLS Bank International has resulted in the invalidation of a huge number of software-related patents over the last 2-3 years (the so-called #AliceStorm), so, depending on when Burke’s patent was issued, it might be made up of just the sort of abstract claims that a court might strike down under Alice. In addition, even if the patent wasn’t swept up in the #AliceStorm, if the claims are as absurd as Richard thinks, they might be interpreted narrowly, such that they might not cover Pied Piper’s product. Finally, how about Ron LaFlamme’s $22,000 invoice for the joint defense group paperwork, which Pied Piper’s business manager describes as “high end of standard?” Let’s see . . . 27.5 hours at $800 per hour? Yeah, that probably checks out. Walking on Sunshine After Burke increases his demand to $300,000 to punish Pied Piper for its defiance (a not-uncommon litigation tactic), it seems like disaster for the funding-starved company. Much like “Two Days of the Condor” and its deus ex restrictive covenant, though, the day is saved once more by an unrelated legal wrinkle. In Richard’s first encounter with the patent troll, Burke discusses his prior legal hustle. He made his fortune buying up the rights to obscure songs and using them to sue the authors of similar-sounding hits. His crown jewel is the mariachi number “Canción de Amores,” which apparently sounds similar to lots of other more popular songs, including Katrina and the Waves’ 1985 TV-commercial staple “Walking on Sunshine,” of which the lawyer now owns a 10% stake. Only when, according to Burke, “all the good music copyrights were gone” did he turn to suing tech startups.
Backed into a corner, Richard gambles on a desperate plan. He dusts off an old program he wrote which was intended to analyze pop songs mathematically. He then confronts Burke with evidence (fabricated as it turns out) that his precious “Canción de Amores” was in fact ripped off the even earlier country tune “I’m Cryin’ (with My Eyes).” Unless Burke drops his claims against Pied Piper, Richard threatens to present this information to Katrina and the Waves, who will no doubt choke off Burke’s royalties.
The patent née copyright troll sneers at first, but Richard impresses on him the persuasive value of “the most advanced and sophisticated pattern matching algorithm ever designed by a genius.” “That’s computer science!” he adds. Burke relents. It’s a stand-and-cheer moment for the show’s core nerd demographic. A victory for the builders with pocket protectors over the leeches in suits. Too bad it’s so implausible. After a track record of carefully researched IP storylines, Silicon Valley finally gives us a legal plot that’s outright preposterous. First, Burke’s scheme is not one I’m familiar with. There is a class of plaintiffs referred to as “copyright trolls,” but the term is usually applied to companies that file mass lawsuits against Internet users accused of pirating content (check out our article on Malibu Media). Buying up copyrights in the hope of suing authors of similar subsequent works seems too speculative and time-consuming to make much economic sense. That said, it’s not unusual for the authors of musical compositions to successfully sue the publishers of later (usually more successful) songs, claiming that they sound too much alike. The most recent high-profile example was the Marvin Gaye estate’s $5.3 million award in their suit against Robin Thicke over “Blurred Lines.” But there is a long precedent for holding the makers of copycat pop songs liable, including the 1976 decision finding that George Harrison’s number-one single “Oh Sweet Lord” was copied from The Chiffons’ “He’s So Fine.” These kinds of claims aren’t easy, though. Copyright infringement by definition involves (1) substantial similarity between the works; and (2) actual copying. While the “Oh Sweet Lord” litigation legitimized the idea of “subconscious plagiarism” (Harrison claimed never to have heard “He’s So Fine” when he wrote his hit gospel tune), you must show that a prior work like “Canción de Amores” had such wide exposure that the subsequent author couldn’t help but be aware of it. In contrast to patent infringement, where ignorance of the invention is no excuse, a totally unknown work forms a poor basis for a copyright claim. On the other hand, based on the rendition in the episode’s closing credits, “Canción de Amores” does sound suspiciously like “Walking on Sunshine.” So maybe Burke just lucked out.
The more fanciful aspect of this plotline is Burke’s reaction to Richard’s algorithm. Leaving aside the fact that Richard’s threat sounds an awful lot like extortion (something LaFlamme correctly counseled him against back in the second season), it just doesn’t have the teeth that the show wants it to have. There are multiple problems. First, Burke’s share of “Walking on Sunshine” is presumably the result of some sort of settlement agreement. Katrina and the Waves can’t walk that agreement back just because they later learn his claims weren’t as strong as they thought (if Burke committed fraud, that’s a different story, but there’s no hint he did). Second, if “Walking on Sunshine” really does infringe on “Canción de Amores,” that infringement isn’t necessarily negated by the fact that both works mutually infringe on “I’m Cryin’ (with My Eyes).”
Most importantly, though, Richard’s faith in the certitude of his algorithm is misplaced. “Substantial similarity” is a fairly subjective standard (frustratingly so for critics of judgments like the “Blurred Lines” case, who point out that all pop songs share certain formulae), and it’s assessed from the perspective of an average audience member, not from the perspective of a music expert, much less the perspective of an AI designed to coldly analyze artwork.
It’s thus not clear that Richard’s algorithm would even be admissible or for what purpose. Expert witnesses can and do testify in copyright infringement cases, and their testimony can arguably sway a jury (especially on the issue of actual copying). However, experienced litigators tend to be confident in their ability to discredit experts. Burke would be foolish to treat Richard’s revelation as a trump card, especially without consulting his own expert.
The show suggests that Burke, in his technical illiteracy, is cowed by Richard’s CS technobabble. The reality would probably be much the opposite. IP litigators are often delighted when they don’t understand what the opposing expert is talking about. It means the judge and jury probably won’t follow it either. So, in the end, Silicon Valley’s writers aren’t above taking a few liberties when it helps bring an intellectual property conflict to a tidy conclusion. Notwithstanding, the show continues to provide a surprisingly detailed and accurate look at the legal issues facing the tech industry.