Recap of the 47th Annual IP Institute

Panelists Elizabeth J. Rest (Partner and Co-Founder at Crown LLP) and David Branfman (Of Counsel for Branfman Mayfield Bustarde Reichenthal LLP) provided commentary on the basics and benefits of licensing and growing one’s brand.

The panelists began discussing how trademark owners may enter into licensing agreements with other individuals and businesses. This agreement entails the trademark owner granting the right(s) to the mark in exchange for royalty payments. This is beneficial to the trademark owner because they may earn income without producing or selling their products/services themselves. Typical licensing agreements include merchandising, settlement of an infringement claim, co-branding, estate planning, and trade dress. One point emphasized was that licensing requires management, continuous monitoring, and quality control. This is imperative for a brand to maintain its integrity and when performed strategically, owners can gain much revenue.

In addition, licensing allows Brand Extension and Co-Branding.  Brand Extension allows the trademark owner to explore other markets, goods, or services, such as “a well-known clothing brand licensing its trademark for the use on accessories or home goods, expanding its reach and revenue potential.” Brand Extension is likely to succeed in a scenario where the trademark is strong. Co-branding, however, allows the trademark owner to collaborate with other companies to grow their customer base; in fact, they can essentially double their market.

Concluding their presentation, both panelists touched on the topic of franchising. It was noted that the licensing process could lead to accidental franchises. As panelist Elizabeth J. Rest noted, franchising has its own set of laws, rules, and regulations. There is licensing involved in the fashion industry as well as the food (for food safety) and wine/alcohol beverage industries. Although there is licensing that takes place within the cannabis industry, the panelists noted there are various conundrums for licensing, due to the fact it is still regarded as a federal illegal narcotic. Overall, both panelists Rest and Branfman maintained positive views on licensing and the many benefits it brings for trademark owners.

Panelists Andrew Klein (Perkins Coie LLP), Christopher B. McKinley (Skadden, Arps, Slate, Meagher & Flom LLP), and Jonah Probell (Intellectual Probell) provided insight into recent patent litigation trends.  The main takeaway from the discussion was that the litigation of patent cases declined by 30% just this year compared to the number of filed cases the year prior.

The panelists reviewed statistics published by the Patent Trial and Appeal Board (PTAB) including the filing of cases—month by month—between the dates of June 1. 2022 and June 30, 2023. The filings remained relatively constant until March 2023, when there was a sharp decline that continued until May 2023. To offer context, 71% of the filed petitions were electrical/computer technology, 18% attributed to mechanical & and business methods, 8% to biology/pharmaceutical, 2% to chemical, and only 1% to designs. One point the panelists emphasized was the use of artificial intelligence and whether there would be more litigation in that specific realm. Nonetheless, litigation regarding artificial intelligence should be monitored.

On the topic of litigation, most patent lawsuits are filed in the Eastern and Western Districts of Texas. The Western District may be the most ideal place to file due to its speediness.  Plus, as noted by the panelists, “Texas values property rights.” Part of the speediness of litigating is due to a ruling by Judge Albright—in hopes of creating efficiency and moving along cases, which is beneficial in some cases but could pose a challenge in others. Because of these challenges, there has been a downturn in filing in the Western District for this reason.

Another trend is enablement challenges under Section 112b. There is concern that the petitions are getting too aggressive with broad claims. Since these claims were filed in the Western District (most commonly), a question arises whether the judges assigned to these cases will adopt rulings like Judge Albright due to his influence within the Western District. Another question posed was will this deter litigants from filing in the Western District. Lastly, will the judges let the jury weigh in on enablement (112b) arguments? This is questionable because generally claim constructions are legal issues and, in practice, are not brought before a jury.

Every day it seems there is a new type of AI and with every new AI comes an onslaught of copyright infringement claims. At a high level the basic premise of these copyright claims is that the training of a model uses content taken from the internet that may include copyrighted work, and if that content is used without authorization, it constitutes direct copyright infringement. Some of the common themes asserted in these copyright infringement cases include:

1. Attacks on unauthorized use of copyrighted works to train AI models
2. Fair Use as an asserted or presumptive defense
3. Allegations that AI models are infringing derivative works
4. Allegations that outputs are infringing derivative works
5. Different views on whether the model provider is directly or secondarily liable for infringing outputs
6. Assertions of violations of the copyright management information provisions of the DMCA
7. Assertions of violations of state laws against unfair competition, negligence, and/or unjust enrichment

In many of these cases, the plaintiffs have asserted that every output must be an infringing derivative work of training material. However, the panelists believe this claim is subject to some vulnerability. Given the nature of these claims, which are pled as class actions, typically plaintiffs have not identified any specific outputs they are challenging. Instead, plaintiffs are trying to attack the model itself and how these models function. However, this is difficult when there is little trace of the training material in the end product.

To help predict how these copyright cases may be decided, panelists Heather Whitney and Angela Dunning looked to the Thomson Reuters v. ROSS Intelligence Inc. case where the District Court of Delaware denied cross-motions for summary judgment on copyright infringement and fair use. In that case, ROSS wanted to create a generative AI model that lawyers could use to ask questions and then receive directly quoted cases. ROSS hired a contractor to scrape the headnotes for them to train their model on how to answer questions. Thomson Reuters claimed that when ROSS made copies of the headnotes in connection with those memos in the process of training their model, ROSS was engaging in infringement of Thomson’s reproduction rights.

One issue in that was whether the headnotes were even copyrightable, which the judge did not want to decide and left for the jury. However, much of the analysis of the panel was dedicated to the fair use defense, specifically Factor 4- the effect of the use on potential markets.

The plaintiff argued that Ross’s product directly competed with Westlaw, so there is harm in the loss of customers. However, the judge disagreed, noting it is important in market harm cases to distinguish between technological changes and actual use of expression. What is being produced did not use the expression at all. The Ross case is distinguishable from Andy Warhol Foundation for the Visual Arts, Inc (“AWF”) v. Goldsmith, because in Warhol, there was a copying of a copyrighted work in its entirety. Here, the Court found that the output is so attenuated and had created something that was entirely new.

A difficulty for Factor 4 of the Fair Use defense is the AI creates a situation in which the very tool that is being challenged is what created the market that Plaintiffs say will be interfered with. The remaining concern for these future copyright infringement cases is: How are we defining the market?

The panel then pivoted to another issue – copyrightability. Can you copyright a work you have created while using these tools? What does it mean to be an original work of authorship? As a threshold matter, we know from cases like Naruto v. Slater, the author must be human to be entitled to copyright protection. But that does not tell us what degree of human authorship is required when using new technology to create art. To answer the question, the panel looked to a very old case that created a standard for copyrightability. In Burrow-Giles Lithographic v. Sarony, the court decided that photographs were entitled to copyright protection so long as they displayed a modicum of creativity. As photographs challenged copyright law in 1884, AI challenges us today.

The internet is often compared to the Wild West. It can be scary and seem like a place with no sheriff. But like the Wild West the internet is not a lawless land, but a land where there is either a problem with the law or a problem with enforcement. According to keynote speaker Eric Goldman of Santa Clara University School of Law, Section 230 forces us to confront this question.

In a nutshell, Section 230 of the Communications Decency act provides that websites are not liable for third-party content. A simple concept, perhaps, but one that opens doors to a multitude of issues.

The Good: (Web 2.0) Not people talking at us but us talking to each other. For how scary the internet can be, we often overlook the good of the internet and how it allows us to communicate with each other. Section 230 allows us to talk to each other. Before Section 230, there was the moderator’s dilemma, either: (1) do nothing, let users have their say, and you wouldn’t be liable for what they said because you weren’t even trying or (2) if you didn’t want to be liable you had to moderate it perfectly by eliminating all the risk in the user content. In comes Section 230 with an answer to the dilemma — to act or not to act, it does not matter because the answer is the same. Moderators shall not be harmed.

Eric Goldman argues that because of Section 230, social media sites can give a voice to marginalized groups and allow people to find a community. Furthermore, it not only creates online marketplaces, but marketplaces where we have more options at lower prices when mobile stores are competing with one another. We also get consumer reviews because it guides the market to make better decisions by rewarding good producers and punishing bad producers. As a result, the market is more responsive to consumers’ needs.

The Bad: The Demise of Section 230. Have we hit the height or limit of social benefits of Section 230 and are now spiraling down under attack from politicians and jurisprudence? The case Eric Goldman uses to highlight this phenomenon is Lemmon v. Snap Inc., where Snapchat’s speed filter allowed users to superimpose the speed at which they were traveling on the content itself. This feature was being used by young adults to drive recklessly, and the courts found Section 230 did not apply because Snap is liable for the content offering tool that encouraged people to engage in unsafe behavior. However, Eric Goldman expressed wariness of the court’s reasoning, namely that “Snap is not eligible for Section 230 protection if it negligibly designed its tools” since this could become a claim that every Plaintiff will argue in every Section 230 case.

Section 230 is under attack from multiple vectors. Social media is investigated as being addictive. Some want Section 230 to be repealed completely or changed in a way rendering Section 230 unrecognizable. Goldman painted a picture of two different internets and the painful trajectory that Section 230 is heading toward.

The Ugly: A world without Section 230. If we add liability or create a regulatory superstructure to control conversation, it shifts the supply curve. We are going to see prices go up and quantity going down, where the supply is our ability to talk to each other. The ugly is people talking at us instead of us talking to each other. But the ultimate question is, who will be allowed to talk?

The panel consisted of Chad Shear (patent litigator), William Raich (patent litigator), and moderator Samantak Ghosh (Intellectual Property Counsel). After a brief introduction, the moderator, Mr. Ghosh, provided an overview of the Safe Harbor provision. He explained how the Hatch-Waxman Act enacted in 1984 played a pivotal role in balancing the incentives of branded and generic drug makers, significantly reducing the time required for FDA drug approval, cutting approval times in half. Mr. Ghosh highlighted the protective nature of the Safe Harbor provision against patent infringement, particularly in activities related to the development and research of new drugs. Over the years, there have been efforts to comprehend the specific statute and its role in the FDA approval process.

Chad Shear delved into the Proveris Scientific Corp v. InnovaSystems, Inc. case, emphasizing its impact on the Safe Harbor provision. He described Safe Harbor as one of the most intricate and challenging areas of law due to its unpredictability and the substantial financial stakes involved in the pharmaceutical industry. The discussion touched on the complexities arising when advising clients, particularly in cases involving research tools. Proveris was pivotal, as one of the first cases involving research tools and whether or not its tools fall within the Safe Harbor provision. The Federal Circuit ruled that Innova’s optical spray analyzers (OSAs) did not qualify for the Safe Harbor, as Proveris’s patented product was not subject to the required FDA approval process, making it ineligible for the benefit of the patent term extension afforded by 35 U.S.C.§156.

The panel further discussed the Regenxbio, Inc. v. Sarepta Therapeutics, Inc. case, where Sarepta utilized patented cells to develop its own product. Regenxbio advocated for the Proveris interpretation of the Safe Harbor. Mr. Raich was personally involved in this case. The panelists also touched on the Allele Biotech v. Pfizer/BioNTech case, where Pfizer’s motion to dismiss was denied as it failed to demonstrate that the patent (and the technology covered by it) constituted a patented invention subject to FDA review.

In the final segment of the panel, the discussion centered around the interpretation of “reasonably related” in the Safe Harbor provision Section 271(e)(1). It does offer broad protection, essentially gauging how reasonably a matter is linked to FDA activities. Mr. Shear explained how the word “solely” in this provision means that certain activity can fall within the protection today but it is not a guarantee for tomorrow. The seminal case that exemplified this is Merck KGaA v. Integra Life Sciences, 545 U.S. 193, 202 (2005), where the court clarified that “reasonably related” does not necessitate direct submission to the FDA. Rather, any research conducted for the purpose of submission qualifies. One of the key concluding points made was that both medical devices and drugs are encompassed within the Safe Harbor, irrespective of whether premarket approval is required.

The panel briefly touched upon the intricacies of assessing damages in this context. They concluded by emphasizing that the Safe Harbor is more unpredictable than merely determining whether a certain device qualifies as a patented invention or not.

Patrick Reilly (US Patent Attorney) commenced this session by introducing the panel and outlining three key topics to be discussed: Obviousness-type double patenting, admissibility of post-filed data, and assignor estoppel. The other panelists were James Snaith (European and UK Patent Attorney) and Marc David Peters (US Patent Litigator).

Marc Peters started the discussion by tracing the evolution of patent terms, noting their increase over the years to the current 20-year patent term. The conversation then delved into the doctrine of obviousness patenting and how patent term extension (PTE) can extend term beyond terminal disclaimer as set in the Merck & Co. v. Hi-Tech Pharmacal Co. case. Mr. Peters also pointed out that patent term adjustment (PTA) cannot extend the term beyond the terminal disclaimer.

Mr. Snaith provided insights into how double patenting differs in European laws. He highlighted that the European Patent Office (EPO) follows the Enlarged Board of Appeal decision, citing Article 125 “In the absence of procedural provisions in this Convention, the European Patent Office shall take into account the principles of procedural law generally recognized in the Contracting States.”

Mr. Snaith acknowledged that the article is broad and vague

and that the EPO does not recognize restrictions on obviousness-type double patenting. However, he suggested that these laws are likely to evolve over the years. Both speakers agreed that in the future, the two patent systems are likely to merge, moving closer than they currently are.

Next topic addressed by Mr. Snaith was the admissibility of post-filed data, highlighting a significant change in the law through the EPO Enlarged Board of Appeal Decision G 2/2. He noted that it stated: “A patent applicant or proprietor may rely upon a technical effect for inventive step if the skilled person, having the common general knowledge in mind, and based on the application as originally filed, would derive said effect as being encompassed by the technical teaching and embodied by the same originally disclosed invention.” Mr. Snaith emphasized that this marks a major departure from the previous law.

The final topic covered was assignor estoppel, discussed by Mr. Peters, who noted that this doctrine has been reaffirmed by the US Supreme Court. The seminal case discussed was Minerva v. Hologic (2021). Mr. Peters elaborated on the three ways the Supreme Court stated when the doctrine would not apply. Mr. Snaith compared it with Europe, noting that this issue does not exist over there.