Executive Summary
In 2025, patent litigation stopped being something companies managed and started being something companies had to architect. Across four continents, a set of rulings, arbitration outcomes and procedural maneuvers quietly rewired how intellectual property gets enforced, valued and weaponized. None of these decisions made global headlines. Together, they changed the rules.
What made 2025 notable was not a single doctrinal shock, but the way multiple forums began to interact as one enforcement system. The Federal Circuit tightened the boundaries of AI patent eligibility.
The Unified Patent Court (UPC) emerged as an increasingly assertive FRAND and SEP venue, including first-of-its-kind anti-interim-license relief. The ITC continued to show strong complainant-side momentum in SEP-heavy investigations. And national courts in Germany and China deepened the globalization of FRAND compliance analysis.
For technical organizations, the implications are now strategic rather than reactive. Patent outcomes increasingly depend on how well companies align product architecture, standards participation, licensing posture and evidentiary discipline across jurisdictions.
The organizations best positioned to succeed are no longer those with the broadest portfolios alone, but those capable of integrating engineering decisions with coordinated global enforcement strategy. That is the thread running through every case in this article.
Table of Contents
AI patents face a harder eligibility reality
The most important AI patent decision of 2025 came from the Federal Circuit’s ruling in Recentive Analytics v. Fox. The court affirmed dismissal under Section 101, holding that the asserted claims were directed to applying generic machine-learning techniques to broadcast scheduling and content optimization rather than to a patent-eligible technological improvement.
The ruling matters because it captures the legal tension surrounding modern AI portfolios. Machine-learning workflows are increasingly seen as general-purpose analytical tools. Merely applying those tools to a new data environment is unlikely to survive eligibility scrutiny unless the patent clearly demonstrates a measurable systems-level improvement.
For engineering teams, this shifts the drafting burden toward demonstrating a concrete technological implementation under Section101: patents must articulate how specific model architectures, data-processing pipelines or system-level integrations produce a measurable improvement in computer functionality or another technical field, rather than merely claiming the use of machine learning to achieve a business or analytical result.
The case reinforces a broader 2025 theme: courts are rewarding evidence of technical implementation, not conceptual breadth.
While U.S. courts were narrowing what AI patents can claim, a new enforcement power was quietly asserting itself across the Atlantic.
Europe’s Rise Accelerates as the UPC Enters the SEP Mainstream
By its second full operating year, the UPC had clearly moved beyond institutional experimentation. Legal reporting indicates that the UPC received 239 new infringement actions in 2025, up from 155 in 2024 which amounts to a 54.2% increase, confirming that Europe is no longer a secondary theatre in global patent enforcement.
That growth has been especially visible in SEP and FRAND litigation, where the UPC’s ability to shape pan-European leverage has become strategically significant.
The most consequential example came in InterDigital’s dispute with Amazon over video-streaming SEPs. In a landmark order, the Mannheim Local Division granted what has been widely described as the world’s first anti-interim-licence injunction, prohibiting Amazon from seeking interim licensing relief in the UK that could effectively restrain InterDigital’s ability to enforce its patents before the UPC.
The importance of the ruling extends far beyond the parties. The court effectively treated an interim RAND license mechanism as the functional equivalent of anti-suit relief where it could undermine European enforcement rights. That marked a major procedural escalation in the growing UK-versus-EU contest over who gets to shape SEP rate-setting timelines.
For implementers, the practical consequence is immediate: forum sequencing in FRAND disputes can now materially affect licensing leverage before any merits determination is reached. In particular, the availability of interim FRAND licences, historically used in some jurisdictions to stabilise exposure while infringement and validity issues are litigated, is no longer a reliable lever if courts are willing to restrain such mechanisms. That shift reduces the ability of defendants to defer commercial resolution while contesting claims, increasing pressure to engage in substantive licensing negotiations earlier in the dispute cycle.
The courtroom, however, was only part of the story. Behind the headlines, arbitration was doing something courts rarely do: it was setting prices.
InterDigital–Samsung shows how arbitration now shapes SEP economics
While the UPC development became the headline judicial event, the InterDigital–Samsung relationship remained commercially significant in 2025 through arbitration-driven rate setting.
Mid-year proceedings reportedly produced a royalty outcome widely understood in the market to exceed $1 billion in value, resetting the economic baseline between one of the industry’s most sophisticated licensors and one of its largest implementers. Subsequent moves toward additional arbitration late in the year showed that even after major rate-setting outcomes, SEP economics remain fluid.
This is increasingly how major FRAND conflicts unfold: courts establish leverage, arbitration defines economics and follow-on proceedings recalibrate future exposure.
For standards-driven businesses, this reinforces a crucial point. SEP risk no longer crystallizes in a single verdict. It evolves through multi-stage dispute architecture spanning courts, arbitral forums and licensing resets.
If arbitration defined the economics, the ITC defined the urgency.
ITC momentum strengthens complainant leverage in SEP-heavy disputes
One of the reviewer’s most important concerns was the absence of ITC coverage. In 2025, that omission would materially understate the enforcement landscape.
The ITC continued to display a noticeable pro-complainant, pro-patentee trend, particularly in technically dense SEP and multimedia investigations. That trend matters because exclusion-order risk often exerts settlement pressure faster than district-court damages.
Two investigations stand out.
Nokia v. Amazon – Inv. Nos. 337-TA-1379 & 1380
Nokia’s ITC actions against Amazon over video-capable electronic devices and streaming products became a major focal point for multimedia SEP enforcement. The investigations reinforced how ITC proceedings can become the most immediate leverage point in platform and device disputes, particularly where import restrictions threaten ecosystem continuity.
Ericsson v. Motorola Mobility — Inv. No. 337-TA-1375
Ericsson’s ITC action strengthened the broader narrative that SEP holders increasingly view the Commission as an acceleration forum for licensing outcomes, especially where 5G and mobile-device ecosystems are involved.
The strategic lesson emerging across forums is clear: exclusion risk now functions as a commercial forcing mechanism, often more powerful than eventual damages. This dynamic, long associated with the ITC, is increasingly visible in Europe as well, where the UPC, alongside German courts, is beginning to mirror similar leverage through the credible threat of rapid injunctive relief. The key distinction, however, lies in cost and scalability. UPC proceedings are generally less resource-intensive than parallel U.S. litigation, making them an attractive first step in enforcement strategy. As a result, complainants may increasingly initiate actions in Europe to establish early leverage, before expanding disputes into higher-cost jurisdictions such as the ITC or U.S. district courts.
Not all of 2025’s ITC developments came from the telecom world. One case quietly expanded who gets to use the commission in the first place.
Domestic Industry Doctrine Evolves Beyond Telecom
A particularly relevant doctrinal development came outside the telecom sector.
Lashify v. ITC became one of the most discussed 2025 Federal Circuit matters involving the meaning of domestic industry “use.” Because the case was effectively brought against the ITC’s interpretation itself, it attracted outsized attention from practitioners assessing how non-manufacturing and hybrid commercial models can satisfy Section 337 requirements.
The significance of Lashify lies in its recalibration of the domestic industry requirement under Section 337. The Federal Circuit clarified that the economic prong is not confined to proof of actual product sales or manufacturing activity in the United States, but can be satisfied through substantial investments in activities that exploit the patented technology, including sales, marketing, warehousing, distribution and quality control.
Crucially, these investments must be tied to products that practice the asserted patents, but they need not involve traditional manufacturing operations. The practical effect is a materially lower evidentiary threshold for establishing domestic industry, particularly for brand-driven, platform-based or non-manufacturing entities. This is likely to expand the class of eligible complainants and, in turn, drive increased reliance on the ITC as an enforcement forum, especially where exclusion-order leverage can be used early in a dispute.
Germany. Then China. VoiceAge EVS showed what happens when a patent holder stops treating jurisdictions as alternatives and starts treating them as a sequence.
VoiceAge EVS turns FRAND into a True Global Compliance Issue
The VoiceAge EVS dispute with HMD is best understood not as a single lawsuit, but as almost a playbook for modern SEP enforcement. It began in Germany, where VoiceAge secured injunctions after courts concluded that HMD had not acted as a willing licensee under FRAND principles. What was notable was not just the outcome, but the court’s approach: rather than rigidly following a step-by-step FRAND checklist, German judges looked at the overall conduct of the parties, asking a more practical question: had the implementer genuinely engaged in good-faith licensing, or was it delaying the inevitable?
From there, the dispute moved east. In China, the same conflict took on a different but complementary dimension. Courts found that HMD had infringed VoiceAge’s EVS patents across a wide range of devices and, crucially, had failed to meet its FRAND obligations. The remedy reflected that shift in emphasis: a conditional injunction that effectively gave HMD a choice — take a licence within a defined window, or face a sales ban. It was less about punishment and more about forcing resolution.
Seen together, these decisions tell a larger story about how SEP enforcement now operates. Germany establishes the credibility of injunction risk. China applies pressure through compliance and conditional relief. And other jurisdictions, including the United States, continue to shape valuation and damages. Each forum plays a role, and each decision builds on the last. For implementers, the effect is cumulative: the space to delay, fragment or strategically prolong disputes is steadily shrinking. FRAND, in this landscape, is no longer just a legal argument. It has become a global standard of conduct, tested simultaneously across courts.
For all the attention on Europe and the ITC, U.S. courts were not standing still.
U.S. verdicts restore balance to the global survey
To ensure the article did not over-rotate toward FRAND procedure, 2025’s U.S. verdict landscape also deserves attention.
Intel v. VLSI
The Intel–VLSI licensing battle remained one of the most commercially significant U.S. patent disputes of the year, with multibillion-dollar stakes shaping broader debates around portfolio valuation, monetisation and the future of large-scale semiconductor licensing.
Masimo v. Apple
The continuing Masimo–Apple watch-tech war added another highly visible example of how product-level leverage, import risk and medical-sensing innovation intersect. Few disputes better captured the convergence of consumer hardware, health-tech IP and platform lock-in.
These cases matter because they restore a critical truth to the 2025 story: Even as FRAND and UPC disputes globalize, U.S. courts remain central to valuation and leverage.
Strip away the case names, the jurisdictions and the dollar figures, and every development in 2025 pointed at the same thing.
The Bigger Lesson - IP Management Now Begins Inside Engineering
The strongest lesson from 2025 is not doctrinal. It is organizational.
Across AI eligibility, UPC FRAND intervention, ITC complainant momentum, codec litigation and semiconductor verdicts, the same principle repeatedly emerged: Disciplined IP management and evidence-driven engineering are now inseparable.
Proper due diligence, claim-to-product mapping, negotiation hygiene, standards participation strategy and litigation sequencing all now begin far earlier than legal filing dates.
In practical terms, patent enforcement has become an engineering-adjacent discipline.
Conclusion - 2025 Was the Year Patent Litigation Became System-Level Strategy
The developments of 2025 suggest that patent enforcement is no longer defined by individual rulings, but by how those rulings interact across jurisdictions. AI eligibility decisions are narrowing what can be protected. SEP disputes are unfolding simultaneously across Europe, China, the ITC and arbitration forums. Procedural tools, from anti-interim-licence injunctions to exclusion orders, are shaping outcomes before merits are even reached.
In this environment, litigation strategy is becoming increasingly structured. For many complainants, Europe — particularly the UPC and German courts — is emerging as a logical starting point, offering relatively faster proceedings, credible injunctive relief and lower cost compared to U.S. litigation. Early wins or pressure in these forums can then be leveraged into parallel actions in higher-impact venues such as the ITC or U.S. courts, where exclusion orders or damages exposure can accelerate commercial resolution.
At the same time, jurisdictions like China are playing an increasingly important role in enforcing FRAND compliance and shaping licensing behavior, adding another layer to multi-forum strategy. The result is a staged enforcement model: establish leverage early, escalate pressure across forums and converge toward a global licensing outcome.
For technology companies, the implication is clear. Patent disputes can no longer be managed as isolated legal events. They must be planned as coordinated campaigns, aligned with product strategy, standards participation and commercial objectives. In that sense, 2025 did not simply produce important cases, it marked the point at which patent litigation became a system-level strategic function.
