Back

Foley Viewpoints

Where Will the Physical AI Patent Wars Begin?

November 10, 2026

[Updated and Expanded from Where Will the Automotive Patent Wars Begin? (2020)]

The next autonomy patent wars may not be confined to cars. As AI moves into drones, robots, defense systems, warehouses, and industrial equipment, the collision between software, sensors, connectivity, and physical control may create new licensing and litigation flashpoints.

Introduction

In 2020, we published “Where Will the Automotive Patent Wars Begin?”, an article examining where the autonomous vehicle patent wars would begin and suggesting that the first disputes were more likely to come from new entrants, cross-border competitors, and licensing negotiations than from traditional automakers. That question now needs to be asked with a much wider lens.

The same convergence of software, sensors, communications, and physical control that turned vehicles into computers on wheels is now turning robots, drones, warehouse systems, and defense platforms into computers that move, sense, and act. Indeed, the first skirmishes of a broader physical AI patent war already appear to be underway, in courtrooms from Delaware and East Texas to Copenhagen, Munich, and Hangzhou.

This article is the fifth in a twelve-week series on autonomous systems, physical AI, and patent strategy. In prior articles, we established an expanded taxonomy of patentable inventions across the physical AI landscape, presented an IP strategy playbook for autonomous systems companies, examined whether patent filings help physical AI companies raise capital, and analyzed what acquirers now diligence in AI and autonomy deals. Those articles focused largely on building patent assets. Here, we turn to the question of how those assets may be leveraged, including in disputes, and where the first battles among operating companies may be fought.

1. The Stakes and the Players Have Changed

1.1 Capital Is Raising the Stakes

Patent activity tends to track investment, and investment in physical AI is substantial. As discussed in our third article, robotics and physical AI startups raised a record $27.6 billion across 1,009 deals in 2025, more than double the $13.7 billion raised in 2024, and physical AI startups raised another $16.3 billion across 492 deals in the first quarter of 2026 alone.[1] Individual valuations have followed: Skild AI raised close to $1.4 billion at a valuation above $14 billion in January 2026, and defense AI company Helsing raised $1.8 billion at an $18 billion valuation in July 2026.[2]

At the same time, the underlying patent landscape has matured. As discussed in our first article, WIPO found that over 1.1 million patent families related to the future of transportation were published between 2000 and 2023, and that five inventor locations (China, Japan, the United States, the Republic of Korea, and Germany) account for over 90% of all inventions.[3] Large, concentrated portfolios held by well-funded competitors in a handful of jurisdictions are conditions under which patent disputes have historically emerged.

1.2 Connectivity Licensing Is Already Contested

Connectivity licensing for connected vehicles has already become a point of friction. Avanci now offers a one-stop 5G Vehicle license covering 5G, 4G, 3G, and 2G standard-essential patents, including C-V2X technology, from more than 90 participating licensors. That program has not avoided conflict. In July 2026, the UK Supreme Court ruled that certain claims brought by Tesla relating to the Avanci 5G Vehicle program should proceed to trial before the High Court. Notably, Avanci emphasized that the Court did not determine whether the license is subject to FRAND or make findings regarding its terms, and the program continues to operate as usual.[4]

The significance for physical AI is straightforward. The cellular, V2X, and wireless connectivity that drones, delivery robots, and fleet-managed machines now depend upon draw on many of the same standardized technologies. It is unclear whether connectivity licensing for robots and drones will follow the vehicle pool model, but the vehicle experience suggests that licensing disputes over connected machines appear increasingly likely.

1.3 Robotics Companies Are Now Suing Each Other

Perhaps the most important development is that operating companies in robotics and adjacent physical AI sectors are now asserting patents against one another, often across borders. The following are some illustrative examples:

As illustrated above, several patterns emerge. First, the disputes span the full physical AI technology stack described in our first article, from locomotion and actuation to robotic operating systems and user interfaces. Second, many involve competitors headquartered in different regions. As The Robot Report observed, in many of these cases companies in different countries have made infringement allegations against one another. Third, the venues have multiplied: the ITC, the Unified Patent Court, German national courts, and Chinese courts are now all active battlegrounds alongside U.S. district courts.

2. Where the Physical AI Patent Wars May Begin

In view of the above, the next wave of physical AI patent disputes appears most likely to emerge along five fault lines:

1. Cross-border competition in industrial and collaborative robotics. The Teradyne disputes suggest that established robot makers are prepared to enforce hardware and software IP rights, including patents and software copyrights, in Europe against newer market entrants. With the UPC offering a single forum covering multiple European markets, companies selling cobots, mobile manipulators, and warehouse robots into Europe should expect this trend to accelerate.
2. Legged and humanoid robotics. The Boston Dynamics and Unitree matters show that locomotion, balance, and actuator designs are already contested. As humanoid robots move toward commercial deployment, the mechanical and control-layer patents that underpin walking, grasping, and balance are poised to become high-value assets and targets.
3. Foundational autonomy software. The Perrone suits illustrate how patents directed to general-purpose robotics software may be asserted against automated driving systems. Patents with claims spanning vehicles, drones, and robots may be asserted across multiple platforms, a dynamic noted in our first article.
4. Connectivity and standards licensing. As with connected vehicles, the integration of cellular, V2X, and wireless technologies into drones and fleet-managed robots may give rise to royalty negotiations and, if those break down, litigation. The Tesla and Avanci dispute offers a preview.
5. Warehouse, logistics, and material handling automation. The OPEX and Fortna matters show that logistics automation, which is capital-intensive and deployed at large scale in customer facilities, has already seen competitor disputes, including at the ITC, where exclusion orders can block imports of accused systems.

Unsurprisingly, these fault lines track the areas of the physical AI market attracting the most capital. As discussed in our third article, defense and security robotics alone attracted $8.0 billion across 234 deals in 2025, and autonomous drones attracted $6.2 billion across 169 deals on a trailing 12-month basis.[12] Where capital concentrates, competitive pressure tends to increase.

3. Strategic Considerations for Physical AI Companies

For founders, in-house patent counsel, and investors, the emergence of competitor-driven patent disputes in physical AI has several practical implications. Companies should consider the following:

• Build a portfolio that holds up under scrutiny. As discussed in our fourth article, buyers now characterize patent assets rather than simply counting them. The same qualities that matter in diligence (claim scope, detectability, enforceability, and clean chain of title) determine whether a portfolio has value in a dispute.
• Prioritize detectable claims. Claims directed to features that can be observed in a competitor’s product, such as locomotion behavior, user interfaces, or system architecture, are far easier to enforce than claims directed to internal algorithms, consistent with the patent versus trade secret framework in our second article.
• File where competitors make and sell. The UPC, German, Chinese, and ITC cases above underscore the importance of the jurisdictional strategy discussed in our second article. Patents in manufacturing hubs and target markets provide leverage that U.S.-only portfolios cannot.
• Treat freedom to operate as a launch milestone. Entrants into Europe, the United States, or China should conduct FTO analyses before major product launches and trade shows, particularly in cobots, legged robots, and warehouse automation.
• Plan for counterclaims and cross-licensing. Many competitor disputes resolve through dismissal or settlement, as the Boston Dynamics and Fortna matters suggest. A deep portfolio provides counter-assertion leverage that can shape those outcomes.
• Expect licensing pressure on connectivity. Companies embedding cellular or V2X connectivity into drones and robots should budget for standards-essential patent licensing and monitor how the vehicle licensing disputes are resolved.

Looking Ahead

The physical AI patent wars are not beginning in a single place; they are beginning simultaneously wherever software, sensors, communications, and physical control converge, and wherever well-funded competitors from different regions meet in the same market. Companies that build deliberate, strategic patent portfolios today will be better positioned to deter, defend, and negotiate when those disputes reach them.

Competitor disputes, however, are only half of the story. Notably, not every plaintiff in the physical AI space is a competitor. As vehicles became computerized, non-practicing entities followed, and the same dynamic is now emerging across drones, robots, and other connected machines. In our next article, we examine how NPEs identify targets in the age of physical AI, what the current litigation data reveals, and what companies building connected, software-defined physical products should do to prepare.

The views expressed in this article are those of the authors and do not necessarily represent the views of Foley & Lardner LLP or its clients. The information provided herein may not be applicable in all situations and should not be acted upon without specific legal advice based on particular situations. Prior results do not guarantee a similar outcome.

[1]PitchBook, Q4 2025 Robotics & Physical AI VC Trends (2025), https://pitchbook.com/news/reports/q4-2025-robotics-physical-ai-vc-trends; PitchBook, Q1 2026 Robotics & Physical AI VC Trends (2026), https://pitchbook.com/news/reports/q1-2026-robotics-physical-ai-vc-trends.

[2]Press Release, Skild AI, Skild AI Raises $1.4B, Now Valued Over $14B (Jan. 14, 2026), https://www.businesswire.com/news/home/20260114335623/en/; Intellizence, Startup Funding Trends in July 2026: AI, Robotics, Energy, & Healthcare (Aug. 5, 2026), https://intellizence.com/insights/startup-funding/startup-funding-trends-july-2026-ai-infrastructure-and-robotics/.

[3]World Intell. Prop. Org., WIPO Technology Trends 2025: Future of Transportation (2025).

[4]Avanci, Avanci Responds to UK Supreme Court Decision in Tesla Litigation (July 27, 2026), https://www.avanci.com/2026/07/27/avanci-responds-to-uk-supreme-court-decision-in-tesla-litigation/.

[5]Boston Dynamics, Inc. v. Ghost Robotics Corp., No. 1:22-cv-01483 (D. Del. dismissed with prejudice Jan. 2, 2025); Boston Dynamics, Inc. v. Ghost Robotics Corp., No. 1:24-cv-00184 (D. Del. dismissed with prejudice Jan. 2, 2025).

[6]HC Robotics v. Int’l Trade Comm’n (Fed. Cir. Aug. 22, 2025); see Steve Brachmann, CAFC Is Unconvinced by Claim Construction Challenges to ITC’s Robotics Patent Infringement Finding, IPWatchdog (Aug. 25, 2025), https://ipwatchdog.com/2025/08/25/claim-construction-challenges-fail-convince-cafc-overturn-usitc-robotics-patent-infringement-finding.

[7]Fortna Sys., Inc. v. Plus One Robotics, Inc., No. 5:24-cv-01274 (W.D. Tex. dismissed with prejudice Dec. 10, 2025).

[8]See, e.g., Complaint, Perrone Robotics, Inc. v. Tesla, Inc., No. 1:25-cv-02156 (E.D. Va. filed Nov. 24, 2025) (subsequently transferred); Complaint, Perrone Robotics Innovations, LLC v. Hyundai Motor Am., Inc., No. 1:25-cv-02180 (E.D. Va. filed Nov. 25, 2025) (subsequently transferred).

[9]Eugene Demaitre, German Court Rules in Favor of Teradyne Robotics, Issues Injunction Against Elite Robots, Robot Rep. (Apr. 21, 2026), https://www.therobotreport.com/german-court-rules-in-favor-of-teradyne-robotics-issues-injunction-against-elite-robots/; Teradyne Launches UPC Patent Case Against JAKA, The Engineer (Sept. 4, 2026), https://www.theengineer.co.uk/content/news/teradyne-launches-patent-case-against-jaka (describing the German action as a copyright infringement case).

[10]Steve Crowe, Teradyne Robotics Sues JAKA over 3 Universal Robots Patents, Robot Rep. (Sept. 10, 2026), https://www.therobotreport.com/teradyne-robotics-sues-jaka-over-3-universal-robots-patents/; Steve Crowe, JAKA Hits Teradyne Robotics Back over Patent Dispute, Robot Rep. (Sept. 3, 2026), https://www.therobotreport.com/jaka-hits-teradyne-robotics-back-over-patent-dispute.

[11]Luweimei v. Unitree Robotics, (2025) Zui Gao Fa Zhi Min Zhong No. 756 (Sup. People’s Ct. 2026) (China).

[12]PitchBook, Drone Deals Fueled VC’s 139% Surge into Defense & Robotics (2026), https://pitchbook.com/news/articles/drone-deals-fueled-vcs-139-surge-into-defense-robotics.

Author(s)

Christopher J. McKenna

[email protected]

Boston 617.342.4057

Chethan K. Srinivasa

[email protected]

Boston 617.342.4018