Ex Parte Reexaminations Poised to Make a Quiet Comeback: Discretionary Denial Guidance for Inter Partes Reexamination May Increase Ex Parte Reexaminations
Key Takeaway: When facing a patent infringement suit, accused infringers traditionally turned to inter partes review (IPR) as a faster, more cost-effective alternative to district court litigation. However, recent guidance from the USPTO suggests that IPRs will become harder to initiate, prompting a renewed interest in ex parte reexaminations (EPRs)—a non-adversarial, lower-threshold procedure that allows any party to challenge a patent's validity through prior art. With rising EPR filings and declining IPR use, EPRs appear poised for a quiet resurgence as patent challengers seek alternatives amid shifting USPTO policies.
Imagine this. You were just served with a Complaint for patent infringement and learn that, some years ago, your competitor was granted a patent giving them a legal monopoly to exclude others, including you, from making, using, or selling a particular invention. Worse, after reviewing, you determine that your latest-and-greatest product is uncomfortably close to infringing the patent. Fortunately, you know that your competitor was not the first to develop that technology and that the Patent Office made a mistake in granting your competitor's patent. The day after receiving the Complaint, you dust off a 30-year-old textbook from college describing exactly what your competitor claims to have invented. You know you have the opportunity to invalidate the patent in the district court case, but you also know that process will be long and costly.[1] Further, you know that it can be challenging to convince a lay jury that the Patent Office made a mistake—particularly under the heightened "clear and convincing" evidence standard in district court.
The Chiplet Shift: Evolving Interface Standards and Commercial Ecosystems
Key Takeaway: For attorneys working in semiconductor transactions, the growing use of chiplets means that interface-level diligence is becoming more relevant. Provisions addressing indemnity, licensing representations, and SEP-related obligations are taking on greater importance in contract negotiations. As chiplet adoption becomes more widespread, aligning agreements with emerging standards and ecosystem expectations will be increasingly important for mitigating legal and commercial risk.
For decades, monolithic system-on-chip ("SoC") designs defined the semiconductor landscape. Introduced in the 1970s1 and refined over several generations, SoCs allowed designers to integrate processors, memory controllers, accelerators, and I/O onto a single die. This design philosophy enabled performance gains, simplified manufacturing logistics, and kept IP ownership relatively centralized.2 That model is now evolving. As chips approach the physical and economic limits of scaling, integration of all components onto a single die (once an advantage) has become a cost bottleneck.3
"Chiplets" offer a modular response. Each chiplet is a smaller, function-specific die (such as a CPU tile, an I/O hub, a signal processing unit, or a memory block), fabricated independently and integrated using advanced packaging technologies.4 This approach allows for increased flexibility and configurability for design of the full chip stack, enabling designers to combine components made using different manufacturing technologies, integrate off-the-shelf building blocks, and scale systems beyond the physical limits of a single chip.5 Moreover, chiplets have proved especially well-suited for artificial intelligence workloads. Chiplets' modular nature allows companies to integrate dedicated AI accelerators, expand memory bandwidth, and scale compute more efficiently, all critical capabilities for training and deploying large models.6 Chiplet adoption has accelerated in parallel.
Trends in Transparency in Patent Litigation Funding
Key Takeaway: Over the past decade, third-party litigation funding (TPLF) has risen to prominence in U.S. patent litigation. The high cost of patent litigation can create an obstacle for cash-strapped investors and non-practicing entities, who increasingly seek out third party funders to fund their patent monetization or enforcement strategies. Read the full article from New York Law Journal.
Our Take on AI: August 2025
EU Issues Code of Practice for General-Purpose AI
On July 10, 2025, the EU published its Code of Practice for General-Purpose AI Models, offering a detailed framework to guide providers in complying with the EU AI Act. The Code focuses on three core pillars: Transparency, Copyright, and Safety & Security. The Copyright chapter is particularly noteworthy for its IP implications, requiring providers to implement a copyright policy, use only lawfully accessible data for training, respect rights reservations (e.g., via robots.txt), and establish safeguards to prevent the generation of infringing content. It also mandates a formal complaint mechanism for rights holders.
Court Rules AI Training as Fair Use in Kadrey v. Meta
In a significant ruling for generative AI development, a U.S. court in Kadrey v. Meta found that using copyrighted books to train the Llama AI model was fair use as a matter of law. In granting summary judgment for Meta, the court reasoned that the plaintiffs failed to adequately demonstrate how this specific use harmed the market for their literary works. While this is a pivotal decision, the court cautioned that this does not create a blanket protection for all AI training, stating that "in many circumstances it will be illegal to copy copyright-protected works to train generative AI models without permission."
State AI Laws Live to See Another Day
The AI state moratorium provision did not survive in the recently passed US tax bill. This provision proposed a 10-year ban on states doing anything to enforce state-specific AI laws. A 99-1 Senate vote officially removed the provision from the bill before it was later signed into law. For now, it appears that states will continue to draft and enforce AI regulations, creating a patchwork of AI regulations for companies to navigate.
AI Counsel Code Podcast
In the episode, "Accelerating AI in Renewable Tech," Maggie Welsh and Michael Silliman discuss the synergistic relationship between AI and clean tech. Michael covers how AI drives both energy demand and solutions for climate change, and why AI innovation is slower in clean tech compared to other sectors. They also delve into challenges such as data interoperability, regulatory complexity, and venture funding. Listen to the full episode here.
July 2025 Intellectual Property Report Recap
In case you missed it, here is a link to our July 2025 Intellectual Property Report.
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.