Santa Clara IP conference: How It’s Going: What Went Wrong?

Moderator: Zahr Said, Santa Clara Law

Mark Lemley (patent), Stanford Law School

After 40 years of radical change, things settled down for normalcy in the last 10 years until Trump. 1980-2017: we grant 350,000 a year up from 50,000; now mostly computer/bio instead of mechanical; most inventors were single and now they’re teams; most inventors were from US and now they’re mostly foreign. University patenting started in 1980; now significant. Now first to file (not first to invent); now 20 years from filing (instead of 17 from grant). Patent thickets; patent continuation practice—multiplication from a single application to multiple patents. Lawsuit numbers have tripled. Jury trial was starting to become a thing in 1970s but a small percentage; now a vast majority of trials are before juries. Product producers used to file against competitors; now 50% of lawsuits are filed by NPEs. Patents were invalidated in 70s at around 65%; now it’s 43-45%. 1982: patent appeals consolidated in Fed Cir. Before 1980 there was no reexamination; then inter partes reexam and post-grant review; IPR proceedings became the way most patent validity disputes were resolved. Introduced district court forum shopping—ED Tex, WD Tex, D Del were not where we litigated in 1980.

Hatch-Waxman/pharma litigation against generics didn’t exist until 1984, so pharma patent litigation grows from there. Patent claim Markman hearings were created in 1990s, cemented in 1996; before then, we didn’t know the answer to “who decides what a patent means, a judge or a jury?”

Then there are substantive changes: patentable subject matter broadened to almost everything, then narrowed again. Major changes in prior art; major changes in interferences; obviousness law changed in fundamental ways, narrowing then broadening. Utility doctrine weakened, written description doctrine and full scope enablement became things. All real disputes about infringement are resolved in Markman hearings; changed law of inducement; new concept of joint or divided infringement; we strike down large swathes of rules on inequitable conduct. Ebay changes the rules for injunctive relief: no longer injunctions as a matter of course. We introduce apportionment of damages (renewed from 19^th century); we change rules on willfullness/advice of counsel and venue.

If you’ve been litigating for 40 years, everything in the system has changed/been in flux, until about 2017. Not much happened since then! A period of normalcy. 62 SCt patent cases 1982-2018, and none in last 3 terms. Recent SCt decisions have had “instinct for the capillary”—clarification of assignor estoppel, not the central issue in patent law. 3 cases on 271(f) about exporting components from the US that are combined outside the US: the only 3 cases on 271(f) of which he is aware. Whether the post office is a “person” under the Patent Act. Not earthshaking! Contrast to KSR, eBay, and patentable subject matter cases before. Substantive cases began to affirm existing law rather than change it. Fed Cir has also settled down: en bancs used to be 2x/any other circuit, but only 2 in last 7 years (1 design patent, another a damages case that was a dud). Fed Cir Dissent rate dropped from 2d highest to one of lowest; many cases now not precedential.

IPRs turned out to have the same invalidity rate as courts at 1/10 the cost. Even patentable subject matter is pretty predictable. There are still cases but they don’t make fundamental changes.

This is generally a good story.

Then Trump. All is in flux. New PTO director dismantling large swaths of PTO, making IPRs essentially impossible to get; increased quotas; refused to hire new career examiners, offers no promotion path; on track to replace examination with AI; Lemley is skeptical. Part of a broader chaos targeting scientific research; 100% tariff on patented pharmaceuticals to somehow magically reduce drug price; taxing university patents; proposing tax on patents’ assessed value at time of filing. Not clear whether normalcy survives.

Mark McKenna (trademark), UCLA School of Law

Conceptual changes that transformed what TM is mostly for the worst. (1) unbounded expansion of the concept of source in TM. Source used to mean actual historical source: who made the thing you were buying. Infringement was very tightly limited to where consumers would believe D’s products came from P. That was passing off. Only someone tricking consumers into buying directly competing goods—very tightly connected to TM’s theory of harm: illegitimate diversion of trade. Courts, primary drivers here, came to regard that definition as overly limited; commerce was changing and courts wanted to capture newer commercial practices, specifically outsourcing of production and expansion of product lines to adjacent areas by companies—wanted to let Coca Cola to license production to independent bottlers and still maintaining rights. Redefined source as “control over quality” instead of actual historical source. Also started to recognize confusion about use of same mark on products that didn’t compete directly, like pancake mix and syrup. Unmoored TM from traditional focus on direct competition and we never replaced it with a real limit. “Sponsorship or affiliation” confusion is the worst; so open-ended that virtually any conceivable connection can be conceived in those terms even though it has different effects on consumers/competition more broadly. Net effect: confusion is performance art—parties need to use that word, but that’s not what the cases are really about, which is rights in gross.

(2) Connected and mutually reinforcing: structural collapse of what used to be a separate but related body of unfair competition law into TM, and resulting expansion of TM subject matter. Once upon a time, only certain things could be TMs, words or devices that didn’t give info about nature of goods/geographic origin: arbitrary/fanciful words/devices that were separate from but attached to the goods. Personal names, descriptive terms, product packaging/product design wouldn’t count; service marks weren’t affixed to anything for sale. Only technical TMs could be infringed; unfair competition law was not about what somebody owned, b/c by definition they didn’t own anything. P had to show that D was passing off even w/o a TM, so there were additional proof requirements and generally much more limited remedies. Would not bar use of descriptive word, etc.

That system collapsed, mostly b/c of Erie. Unfair competition became understood as state law, not common law, and federal courts started to believe they couldn’t have common law, making lawyers worry about 50 different kinds of unfair competition law. That lack of uniformity wasn’t really happening but courts and lawyers worried that it might. Solved by recharacterizing things that used to be excluded from TM and calling them “unregistered TMs.” Those things were previously unregistrable, not just unregistered. Courts started interpreting 43(a) to give a cause of action for infringement of unregistered TMs. Not what Congress anticipated. Huge change. We lost the orientation of unfair competition as a residual doctrine w/more limited remedies and got a huge amount of subject matter where we didn’t have rules about what could be owned. Had to build that law about what could be owned from scratch and haven’t been particularly successful.

These two things amplify each other: TM is redefined as anything that can indicate source as we’ve expanded source beyond recognition. Incremental; radicalism not notice. Courts act like the assumptions they used to hold under the old system still apply, even though the changes are too great to make that true.

Trevor Reed (copyright), UC Irvine School of Law

Indigenous creative rights: © is silent on status of citizens of over 500 tribes, despite importance of Indian culture to 1970s American culture. Many pieces of indigenous self-determination legislation in the 70s.

Mismatch b/t © and tribal sovereignty: tribes often regulate traditional knowledge; © might call it public domain or give it only thin protection. Eurocentric assumptions: limited times, transfers to certain people only, money is a sufficient remedy, it doesn’t matter where the creativity occurs; divides intangible from tangible. Ignores tribal sovereignty; 301 preemption can conflict. Tribes should, among other things, be able to take over registration and deposit. His objections: Should be no public domain for tribal creativity w/o tribal authorization; federal remedies for violation of tribal rights; tribal courts should be recognized as venues for © claims and tribes’ regulatory authority should be formally recognized.

Pamela Samuelson (copyright), UC Berkeley Law

Statutory damages are the worst! Evolution: until 1909 Act, there was a per sheet penalty dating back to the Statute of Anne, and statutory damages were an improvement (anybody could ask for the PSP and half of the money went to the US gov’t; not used often). Particularly useful when difficult/expensive to prove damages; courts had discretion to grant statutory damages but generally wouldn’t if damages or profits were measurable. Nonpunitive way to get some compensation and deterrence.

1976: good parts: tripartite structure of $750-30,000 as court considers just; up to $150K for willfulness; discretion to reduce awards if innocent infringers or nonprofit educ/library users who thought uses were fair. Compensatory at low end; deterrent in middle; punitive at high end.

Understandable but contribute to problems: Ps can ask for SD at any time before final judgment; they’re mandatory.

In practice: Congress failed to consider how they should be assessed in secondary liability or multiple work cases—in Cox, jury awarded nearly 100K per work, $1 billion. Authors Guild v. Google, estimated risk was above $350 billion. Arbitrary, inconsistent, and grossly excessive awards.

Suggested guidelines: award minimum where there’s no actual damages or profits or P is unwilling to provide evidence; approximate actual damages when fair use/lack of infringement is plausible, 2-3x actual/profits when reckless or intentional; up to 10x if highly willful. Or consider what will deter this D. Cox v. Sony: main issue is standard for contributory infringement, but second issue was the standard for willful infringement, but oral argument ignored it. SG and Cox said focus should be whether Cox knew its own conduct was unlawful; reasonable for Cox to think it was OK to continue to provide service to accounts whose users infringed, especially to hospitals and barracks and the like.

Now getting worse: 1202 statutory damages. Very similar but minimum is $2500 w/maximum 25K. Measured per violation (not defined) not per infringed work. No “as the court considers just” limit. No need to have registered © for eligibility. Way more likely to result in excessive damages.

Said: heard a lot about Erie, more than any other IP conference: what gives?

McKenna: Congress has left the field and SCOTUS justices are no longer picked for being lawyers but for specific hot-button issues. That leaves everything to lower courts.

Lemley: we’ve also decided to abandon the common law and equity for the dictionary definition of whatever terms the judge decides to look up, which is a particular disaster for an IP regime that assumed a common-law development. Many key concepts (infringement standard) aren’t even in the statute. So we’ve abandoned the tools we’ve used for centuries. Leaving us with the executive branch, and leaving aside Trump chaos, one of the challenges is that you won’t get long term consistent development. We’re looking at traditional sources of federal law and finding them wanting.

Samuelson: tech is also a big driver in ©. Generative AI, billions of dollars at stake. Every other tech has pissed off a specific sector: recording industry, movies, etc. Now everyone is mad. Dismantling of federal science community is also hurting. Copyright is the only law on the books is the only thing that seems like it could destroy AI; that’s not going to happen but Ps can ask for impoundment/destruction, or they could end up having control over the models.

Reed: people are losing faith in economic rationales of IP; social justice is becoming a bigger rationale and people want to see more of that, but we keep spinning out more economic arguments. Compare backlash to racist mascots—pressure on corporations to change their TMs.

Farley: what went wrong in Dastar? Unbounded definition needed constraint.

McKenna: Dastar’s biggest fan! But courts haven’t applied it. The case is hard: you have to dig in to get what the court is saying. Whatever else you thought of Scalia, he was smart and engaged with the arguments, and that level of engagement is less common. Also, lower courts don’t like the outcomes it produces and so ignore it. That means TM is used as a back door for copyright, especially for works in public domain.