A New World: Indemnification, Artificial Intelligence and Intellectual Property
To describe the rise of artificial intelligence (AI) as rapid is a vast understatement. From its ubiquitous presence assisting us in our most mundane daily tasks to its warm embrace by business leaders eager to revolutionize their industries, AI is pervasive – and so are its pitfalls. Indemnification provisions and enforcement are expected to shift significantly as organizations adapt to related legal changes for creators, manufacturers, distributors and users.
AI and Machine Learning
The integration of AI and machine learning in various industries has led to a new set of challenges for businesses. AI algorithms are becoming more sophisticated, increasing the risk of unauthorized access. We can anticipate to see an uptick in lawsuits related to AI and machine learning as businesses seek to protect their IP and push liability to other entities along the AI spectrum. Additionally, we have seen and will continue to see an uptick in AI policies and procedures and AI risk-shifting contractual provisions to limit risk and mitigate exposure.
Expanded Risk-Shifting with AI
Innovation continues to accelerate. Businesses are constantly developing new technologies and processes with the assistance of AI. Open questions remain as to who should bear the risk when AI is involved.
You used an AI platform to create a specific solution for a customer, and now the customer is being sued for copyright infringement by another party – and the customer is turning to you. Who is responsible? You? The customer? The AI platform? Who has indemnity in this scenario? We work with clients at every stage of product development to address such issues and mitigate risk before it reaches a critical point.
BakerHostetler's Approach to Indemnification with AI
We help clients in all aspects of the supply chain and vendor relationships understand, evaluate and allocate risk across the supply chain, with best practices regarding indemnification and AI management and with contract drafting, interpretation and enforcement. We focus on our clients' business needs and goals, our clients' risk tolerance and industry specifics to craft solutions and provide recommendations.
We handle all aspects of the indemnification life cycle. We not only draft indemnity terms and provide IP strategy relating to AI, but we also enforce these provisions and recover for our clients. Importantly, we have experience allocating indemnity obligations in cases where there are multiple indemnitors, as is often the case with our technology clients. Clients turn to BakerHostetler for legal strategy, risk evaluation and management, efficiency, and our exceptional success in achieving positive results for our clients and recouping money that would otherwise have been lost without sound indemnification policies and practices.
Slings and Arrows: Trade Secret Trends Reflect a Need for Safeguards
Consider trade secret enforcement your armor on the battlefields of IP. Given greater exposure to cybersecurity breaches and data theft, a murky future for employee noncompete agreements, AI algorithms that challenge the integrity of trade secrets – trends suggest businesses will require robust and proactive IP strategies, including an increase in trade secret litigation and internal protection, compliance, and more.
Trade secret trends are expected to undergo a significant shift as organizations adapt to expanded trade secret protection, a potential ban on noncompetes, cybersecurity breaches and data threats, new technologies, and changes in the legal landscape.
Expanded Scope of Protected Trade Secrets
Businesses are constantly developing new technologies and processes that qualify as trade secrets. Innovation continues to accelerate. Patent protection is increasingly more difficult and expensive. High trial success rates, large damages awards and higher probabilities of obtaining injunctions often make trade secret cases a more attractive option than patent cases. Litigation funders are also more likely to back a trade secret case than a patent case. Businesses are increasingly turning to documenting and safeguarding their trade secrets rather than relying on patents. This expansion should result in a continued increase in trade secret litigation and internal protection, compliance, and safeguarding practices, as organizations seek to protect their intellectual property.
Impact of the Future of Noncompetes
Earlier this year, the Federal Trade Commission (FTC) issued a new rule banning noncompetes for most employees. The new rule was supposed to go into effect as of September 4. On August 20, a district court issued an order blocking the FTC from enforcing the rule. According to the FTC, the decision "does not prevent the FTC from addressing noncompetes through case-by-case enforcement actions." If the FTC ban on noncompetes is ultimately enforceable, confidentiality and indemnification provisions in agreements, narrower nonsolicitation provisions, and trade secret litigation will play an escalating role in the protection of a company's intellectual property when faced with employees leaving for competitors.
Organizations also will need to increasingly rely on nondisclosure agreements (NDAs) to protect their intellectual property. If the noncompete ban is deemed enforceable, we can also expect to see an increase in trade secret litigation related to the enforcement of NDAs, confidentiality and indemnification provisions, and broader enforcement of internal protections as businesses seek to hold parties and former employees accountable for breaching the terms of these agreements. Companies must increase their focus on such alternative methods to protect their intellectual property.
Rise in Cybersecurity Breaches and Data Theft
Reliance on digital platforms and cloud storage has exposed companies to increasing cybersecurity breaches and data theft. We can expect to see a continued expansion of trade secret litigation involving cybersecurity breaches and data theft and an increase in contractual indemnification, insurance and warranty protections in NDAs and ancillary agreements, as companies look for means to defend their trade secrets against cybercriminals and mitigate financial exposure.
Artificial Intelligence and Machine Learning
The integration of AI and machine learning in various industries has led to a new set of challenges for businesses when it comes to protecting trade secrets. AI algorithms are becoming more sophisticated, increasing the risk of unauthorized access and misuse of trade secrets. We can expect to see an uptick in lawsuits related to AI and machine learning as businesses seek to protect their intellectual property from misappropriation. Additionally, we have seen and will continue to see an uptick in AI policies and procedures and AI risk-shifting contractual provisions to limit risk and mitigate exposure.
Globalization and Cross-Border Trade Secret Disputes
We can further expect to see a rise in international trade secret litigation as the world becomes increasingly interconnected. Businesses must navigate the complexities of intellectual property protection in multiple jurisdictions. While a number of district courts have allowed extraterritorial application of the Defend Trade Secrets Act against companies acting from outside the U.S., businesses must be prepared to adapt their strategies to comply with international laws and regulations.
As businesses adapt to expanded trade secret protection, the potential loss of enforceable noncompetes, the rise in cybersecurity threats and data threats, increasingly sophisticated AI and machine learning, and globalization, they will need to be proactive in protecting their intellectual property rights. Trade secret enforcement will be an increasingly important component of an organization's intellectual property strategies.
The Changing Landscape: Admissibility of Experts in Patent Cases
Experts play a crucial role in patent cases. Experts opine on claim construction, infringement, invalidity and the proper amount of damages. And the exclusion of an expert witness can significantly impact the outcome of a case. But the standard for excluding experts in patent cases appears to be in flux.
Federal Rule of Evidence 702 – the rule governing admission of expert testimony – was amended in December 2023. The amendment heightened the standard for admissibility of expert testimony. The amendment places the burden on the proponent of the expert testimony to show that it is more likely than not that the opinions meet the requirements of the rule. In addition, the rule now requires that "the expert's opinion reflects a reliable application of the principles and methods to the facts of the case." The amendment puts more power in the hands of a district court judge to potentially exclude expert testimony. And those decisions are reviewed on appeal under the highly deferential abuse of discretion standard.
A few of these cases have made their way to the United States Court of Appeals for the Federal Circuit, which is the appellate court that hears all patent appeals. In 2024, there were three decisions where the Federal Circuit began to explore how the amended Rule 702 applies to experts in patent cases. But in a surprising turn, the Federal Circuit has appeared to lower, rather than raise, the bar on the standard for admissibility of expert opinions.
Reliability
In ParkerVision v. Qualcomm Inc., 1 the District Court excluded the plaintiff's infringement experts, deeming their testimony unreliable due to a lack of necessary testing and simulations to demonstrate infringement, which was recommended by scientific literature. As a result, the District Court granted summary judgment of noninfringement since the defendant's expert's noninfringement opinion stood unrebutted. However, upon appeal, the Federal Circuit reversed this decision, stating that the District Court abused its discretion in interpreting the scientific literature's general statements as a prerequisite for a reliable infringement opinion in this specific case. The Federal Circuit emphasized that it should be up to the jurors to assess the correctness of the underlying facts of the opinion, thereby potentially lowering the standard for the admissibility of expert opinions.
Qualifications
In Osseo Imaging, LLC v. Planmeca USA Inc.,2 the Federal Circuit addressed the qualifications necessary for providing expert testimony from the perspective of a person of ordinary skill in the art. The case involved a patent where the required expertise included three to five years of diagnostic imaging experience as of 1999, the patent's date of invention. The defendant argued that the plaintiff's infringement expert did not acquire the necessary experience until nearly 10 years after 1999, thus rendering the expert unqualified to testify. However, the Federal Circuit rejected this argument, emphasizing that an expert's subsequent acquisition of the necessary skill does not automatically render their testimony unreliable. The court explained that an expert could later develop the requisite understanding and knowledge of what a person of ordinary skill knew at the time of the invention. This ruling effectively broadens the pool of potential technical experts in patent cases, allowing those who have gained the necessary experience post-invention date to still qualify as experts.
Damages
In EcoFactor, Inc. v. Google LLC,3 the District Court refused to exclude the plaintiff's damages expert. The defendant claimed the expert's opinion should be excluded because the royalty rate was conjured out of thin air. Although the royalty rate appeared in three license agreements, these agreements were lump sum deals, and the royalty rate was mentioned only in a "whereas" clause stating the plaintiff's unilateral belief that the lump sum was based on the royalty rate. Despite the plaintiff's lack of access to the sales data of its licensees, these clauses and the testimony of the plaintiff's CEO regarding the royalty rate sufficed for the Federal Circuit to rule that the District Court did not abuse its discretion in admitting the expert testimony.
The defendant also argued that the expert's opinion should be excluded due to the lack of economic comparability of the three license agreements. These agreements licensed multiple patents, not just the single patent in question. The Federal Circuit determined that the District Court did not abuse its discretion, as the expert acknowledged the existence of other licensed patents, which would exert downward pressure on the royalty rate, while the fact that the license agreements were settlement agreements – reflecting the risk of a noninfringement or invalidity finding – would exert upward pressure on the royalty rate.
In allowing these opinions to go to the jury, the Federal Circuit emphasized that "[i]f the standard for admissibility is raised too high, then the trial judge no longer acts as a gatekeeper but assumes the role of the jury." This approach again appears to lower the bar for the admissibility of expert opinions, making it the jury's responsibility to decide whether such testimony is persuasive.
Two important caveats should be noted. First, there was a strong dissent accusing the majority of abdicating the court's "responsibility as a gatekeeper," insisting that the court "must pay close attention to the reliability of the methodology underlying expert testimony to ensure that the jury can fulfill its proper role as the fact-finder."
Second, on September 25, 2024, the Federal Circuit vacated this opinion4 and ordered a rehearing en banc to review "the district court's adherence to Federal Rule of Evidence 702 and Daubert ... in its allowance of testimony from EcoFactor's damages expert assessing a per-unit royalty rate to the three licenses." A decision from the full Federal Circuit is expected in 2025.
Conclusion
While the recent amendment to Rule 702 has heightened the standard for the admissibility of expert testimony, the Federal Circuit's latest decisions reflect a potentially more lenient approach. This approach allows jurors to evaluate the reliability of expert opinions rather than excluding them outright. However, this flexibility might be reconsidered in 2025 when the Federal Circuit revisits Rule 702 en banc in the EcoFactor v. Google case. The forthcoming decision will likely impact the admissibility standards for expert testimony in future patent cases.
On the Rise: Patent Litigation Sees Modest Recovery in 2024
After patent litigation reached a decadelong low in 2023, caseload filings have seen an uptick in 2024, adding to the dizzying ebb and flow experienced in U.S. district courts over the past 10 years. BakerHostetler attorneys have the insight and experience necessary to analyze the numbers, illuminate what the data reveals and explain what likely lies ahead.
Federal District Court Patent Cases Rebound
Patent litigation is experiencing a resurgence. After reaching a decade-low of 3,117 new cases in 2023, the number of filings rose significantly in 2024. New patent cases increased by 20 percent, reaching a total of 3,792, signaling a potential stabilization in patent litigation activity moving forward. The following sections outline the top three patent venues in 2024.
Third Place: District of Delaware
activity in 2024. However, recent data shows that new patent cases in the DDE continued to decrease in 2024, just as they have over the past four years. This reduction to just 395 new case filings can be attributed to growing scrutiny by the court of litigation funding disclosure requirements. Indeed, Judge Colm Connolly of the DDE has continued his quest to ensure that real parties in interest are identified in patent suits. Last year, he sanctioned an owner of asserted patents for refusing to testify before him in person about the financing of the suit. This sanction was affirmed by the Federal Circuit. Thus, the decrease in new suits in the DDE seems driven, in part, by a reluctance of plaintiffs or their funders to expose the nature of their ownership or investment relationship
Second Place: Western District of Texas
The Western District of Texas (WDTX) has become an important venue for patent litigation in recent years, as Judge Alan Albright of the Waco Division has established his court as friendly to patent suits. Although he saw close to 1,000 new case filings in 2021, that trend reversed beginning in July 2022, when the district issued an order mandating that all patent cases filed in the Waco Division before Judge Albright could be randomly assigned to any of a dozen judges within the broader WDTX. Since that order, new case filings continue to decrease because patent plaintiffs are no longer guaranteed to proceed in front of Judge Albright in Waco. New patent case filings in 2024 dropped to just 371, down 62 percent from the high of 989 new cases in 2021.
First Place: Eastern District of Texas
As one of the top patent venues year after year, the Eastern District of Texas (EDTX) continued its prominence as one of the busiest districts in the country for patent litigation in 2024. The number of new patent cases filed in the EDTX rose by over 70 percent in 2024 relative to 2023. The continued rise in filings is likely attributable to the EDTX's perceived plaintiff-friendly environment and efficient case handling, as well as a movement of cases outside other venues as discussed above. Specifically, since patent holders are no longer assured of getting their cases in front of Judge Albright when filing in the WDTX or do not like the new litigation funding disclosure requirements in the DDE, many patent owners have chosen to file new cases in the Marshall Division of the EDTX. Filing here gives plaintiffs about a 90 percent chance of proceeding in front of Judge Rodney Gilstrap, with his pro-plaintiff standing orders.
Conclusion
Making Moves: The USPTO Provides Guidance, Proposes Changes
The U.S. Patent and Trademark Office (USPTO) had a busy summer in 2024. Businesses and individuals with AIand software-based inventions paid particularly close attention to the agency when, in July, it released new guidance on how patent examiners should evaluate related claims for patent subject matter eligibility. The announcement came less than two months after the USPTO solicited public feedback on a proposed new rule for filing terminal disclaimers to overcome obviousness-type double patenting rejections – a change that would prompt patent applicants to take new measures for pursuing continuation applications.
With respect to AI- and software-based inventions, the Guidance describes Prong One of Step 2A (which is renamed from Step 1 of the Alice/Mayo test) as a determination of whether the claim recites an "abstract idea." The Guidance provides useful information around qualifying mental processes – a subcategory of abstract ideas – in particular, that "claims do not recite a mental process when they contain limitations that cannot practically be performed in the human mind[;] for instance[,] when the human mind is not equipped to perform the claim limitations" (emphasis added). If the claim is not directed to an abstract idea, the claim is directed to eligible subject matter.
claim is directed to eligible subject matter. Conversely, the Guidance proceeds to Prong Two of Step 2A if the claim in question is directed to an abstract idea. Prong Two inquires whether the claim recites additional elements that integrate the abstract idea into a practical application, i.e., the "practical application" test. If the claimed subject matter meets the practical application test, the subject matter qualifies as eligible subject matter. If the subject matter does not meet the practical application test, the Guidance proceeds to Step 2B (i.e., Alice/Mayo Step 2) to evaluate whether the claim recites "significantly more" than the abstract idea.
The Guidance introduces three new helpful AI SME examples: Example 47, "Anomaly Detection"; Example 48, "Speech Separation"; and Example 49, "Fibrosis Treatment." The Guidance provides useful insights on how to draft patent applications and claims with an eye toward SME primarily based on the practical application test under Step 2A, Prong Two. Examples 47-49 rely heavily on a specification's disclosure to determine whether any of the claim elements provide a technical solution to a technical problem.
Drafting robust patent specifications is thus more essential than ever in providing the details covering implementation, customization and training of AI/machine learning (ML) models. A specification should provide details on how the AI/ML technology is applied in practice or how the AI/ML technology is integrated with another system or apparatus. Reciting the technical problem confronted in the art and how the invention overcomes the problem helps further support SME. Lastly, at least some claims in an application should be directed to a specific technical solution addressing a technical problem as described in the specification to substantiate SME.
Proposed Changes to Terminal Disclaimer Practice
In 2024, the USPTO solicited public feedback regarding a proposed rule that would limit the enforceability of patents tied together by terminal disclaimers if rendered invalid. Under the proposed rule, if even one claim of a first patent were held invalid, a patent tied to that first patent via a terminal disclaimer could then be unenforceable. The proposed rule has been seen by some in the patent community as posing a threat to continuation applications.
Although the USPTO ultimately withdrew this proposed rule, it is of course possible that some of the issues the proposed rule sought to address – such as concerns over large patent families – may resurface in future proposals.
A New Test: Landmark Decision Overrules Framework for Design Patent Obviousness
The legal landscape quaked, and clients and counsel continue to navigate the tremors. More than 40 years of precedent was upended in May 2024 when a federal circuit court struck down the Rosen-Durling test for assessing obviousness of design patents, sending BakerHostetler attorneys into overdrive as they guide patent owners and challengers in how to obtain and enforce design patents amid the LKQ decision's more-flexible obviousness analysis.
USPTO reports indicate that in the past decade U.S. design patent filings have steadily increased to well over 50,000 filings per year. This increase in filings has likely been fueled, at least in part, by applications focusing on graphical user interfaces and internet-enabled devices.
For 2024, likely the biggest legal issue for design patents relates to the impact of the landmark en banc Federal Circuit decision LKQ Corporation v. GM Global Technology Operations LLC (LKQ Decision). The LKQ Decision has far-reaching implications for both design patent prosecution and design patent enforcement, and accordingly, our attorneys have been assisting clients in navigating these implications..
LKQ Corporation v. GM Global Technology Operations LLC. On May 21, 2024, the en banc Federal Circuit decision in LKQ Corporation v. GM Global Technology Operations LLC overruled the long-standing Rosen-Durling test for deciding whether a design patent was invalid as obvious.
In particular, the Federal Circuit adopted an approach that was consistent with Congress' statutory scheme for design patents, which provides that the same conditions for patentability that apply to utility patents apply to design patents, as well as Supreme Court precedent that suggests a more flexible approach than the Rosen-Durling test for determining nonobviousness.
The Rosen-Durling test requirements were that the primary reference must be "basically the same" as the challenged design claim. Further, this test required that any secondary references must be "so related" to the primary reference that features in one would suggest application of those features to the other. In this regard, the Federal Circuit indicated that the Rosen-Durling test does not adequately align with precedent in terms of its framework and threshold rigidity.
In overruling the Rosen-Durling test, the Federal Circuit referenced the statutory language of 35 USC 103 and the meaning of the statute as addressed by the Supreme Court in Graham v. John Deere Co. of Kansas City, which involved utility patents. Utilizing the language of § 103, as well as precedent on obviousness in both design and utility patent contexts as a guide, the Federal Circuit held that invalidity based on obviousness of a patented design can be determined utilizing the analytical tools for reviewing the validity of a utility patent under § 103, which included application of the Graham factors. The Federal Circuit further indicated that the test associated with the Graham factors has proven workable for utility patents and should be similarly workable for design patents.
The Federal Circuit thereafter addressed in detail how each of the Graham factors should be applied to design patents. These Graham factors as applied to design patents would, at a high level, include:
- Scope and content of the prior art
- Differences between the prior art and the claimed design
- Knowledge of an ordinary designer in the relevant field
- Secondary considerations, such as commercial success, industry praise and copying by competitors
- A Differences between the prior art designs and the design claim at issue
Make It Plain: Clarity Regarding Obviousness-Type Double Patenting
In a year defined by landmark decisions, impactful announcements and new standards, clarity in the patent world comes as a welcome relief. It arrived via a federal circuit court decision in August 2024 that settled certain facts related to obviousness-type double patenting (ODP), though BakerHostetler attorneys warn patentees to remain cautious.
It is very common for patentees to have multiple granted patents within the same patent family. And due to patent term adjustment (PTA), one or more of those patents could expire well beyond the expiration date of other patents in the family. But what if an earlier-filed, later-expiring patent (e.g., a "parent" patent with PTA) could be invalidated simply because there are later-filed, earlier-expiring "child" patents in the family? That has been a growing concern in the patent community over the past few years.
In In re Cellect, for example, the U.S. Court of Appeals for the Federal Circuit concluded that the PTAB did not err in finding that the claims of several earlier-filed, later-expiring patents were unpatentable on the grounds of ODP in view of a later-filed, earlier-expiring patent. That case arose from reexamination proceedings in which the examiner rejected certain claims of Cellect's '621 patent, '626 patent, '742 patent and '369 patent on the grounds of ODP for being unpatentable over the claims of Cellect's '036 patent. The relationship between these patents is shown to the left.
Cellect's '036 patent was the only family member that did not receive PTA and thus expired earlier than its family members.
On appeal, the Federal Circuit held that ODP "for a patent that has received PTA, regardless [of] whether or not a terminal disclaimer is required or has been filed, must be based on the expiration date of the patent after PTA has been added." In re Cellect, LLC, 81 F.4th 1216, 1229 (Fed. Cir. 2023). As a result, the courtconcluded that the PTAB did not err in finding that the earlier-filed, later-expiring patents were invalid under ODP.
This decision sent shockwaves through the patent community and had many patentees scrambling to determine whether their patents, the terms of which had been extended by PTA, were safe from an ODP attack. But on August 13, the Federal Circuit provided some clarity.
In Allergan USA, Inc. v. MSN Lab'ys Priv. Ltd., No. 2024-1061, 2024 WL 3763599 (Fed. Cir. Aug. 13, 2024), the Federal Circuit stated that its holding in In re Cellect relates to the expiration date of a patent rather than the circumstances under which a claim can properly serve as an ODP reference:
Our holding in Cellect is only controlling in this case to the extent that it requires us to consider, in our ODP analysis, the '356 patent's June 24, 2026 expiration date (i.e., the expiration date after the addition of PTA) ...[.] It does not follow, however, that the '356 patent must be invalidated by the '011 and '709 reference patents simply because it expires later. Indeed, Cellect does not address, let alone resolve, any variation of the question presented here – namely, under what circumstances can a claim properly serve as an ODP reference – and therefore has little to say on the precise issue before us.
(Id. at *6; emphasis in original). And in a footnote, the Federal Circuit stated that "Cellect cannot be read as 'confirming,' much less holding, that a later-filed, laterissued, earlier-expiring claim is a proper ODP reference against a first-filed, first-issued, later-expiring claim having a common priority date." (Id.)
In Allergan, the court evaluated whether claim 40 of the '356 patent, the expiration date of which had been extended by 467 days due to PTA, was invalid on the grounds of ODP over certain claims from its "child" patents (the '011 patent and the '709 patent, referred to collectively herein as "the reference patents"). The relationship between the '356 patent and the reference patents is shown above.
The court found that the claims of the reference patents are not proper ODP references that can be used to invalidate claim 40 of the '356 patent. (Id.) In reaching its decision, the court focused on the fact that the '356 patent was the first patent to cover the claimed subject matter. The court noted that it is the first-filed, first-issued patent in a family that sets the maximum period of exclusivity for the claimed subject matter and any patentably indistinct variants. (Id.) Therefore, the court held that "a first-filed, first-issued, later-expiring claim cannot be invalidated by a later-filed, later-issued, earlier-expiring reference claim having a common priority date." (Id. at *7.)
This decision appears to settle the ODP issue only as it relates to the specific facts in this case, given that the court stated that its holding was specific to those facts. Thus, patentees should still be cautious of the ODP implications when PTA is awarded to a patent that is not the first-filed, first-issued patent or when the patents involved are from different families.
*A version of this article was originally published by BakerHostetler in The Patent Lawyer
Footnotes
1. 116 F.4th 1345 (Fed. Cir. 2024).
2. 116 F.4th 1335 (Fed. Cir. 2024).
3. 104 F.4th 243 (Fed. Cir. 2024).
4. 115 F.4th 1380 (Fed. Cir. 2024).
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
