Global patent strategy is shifting. The changes are not all happening at the same speed, and they won’t affect every portfolio in the same way, but they do raise practical questions IP leaders should be asking in relation to capturing, protecting and enforcing their patents over the next few years.
The common thread is that patent strategy is becoming more contextual. Who or what made the invention, how it was made, where it may be manufactured, where it may be enforced, and whether governments regard the technology as strategic all now matter more than they did five years ago.
If your IP strategy was last meaningfully reviewed before these shifts took hold, it is worth revisiting the assumptions behind it.
1. Generative AI is changing how inventions are created
Often patent discussions about AI focus on whether AI-related inventions are patentable. Although that question is important, it is not the only interesting one. Another relevant question for many organisations is whether generative AI is already being used by inventors during conception before anyone in the legal or IP team knows there is an invention to protect.
McKinsey’s 2025 State of AI survey reported that 88% of organisations are now using AI in at least one business function. More importantly for IP teams, 64% of respondents said AI is enabling innovation and 62% said their organisations are at least experimenting with AI agents. In other words, AI is already influencing how work is done, including inventive and confidential work, even though many organisations are still working out how to regulate it properly.
If inventors are using generative AI during ideation, problem-solving or optimisation, the IP team needs to understand what role the tool played. Was it used merely to summarise known information, or did it contribute to the technical concept? Did the inventors input confidential information into a public tool? Were the prompts and outputs saved? These issues can affect inventorship, entitlement, confidentiality, novelty, obviousness and later evidentiary questions. They may also be very difficult to untangle years later if no records were kept at the time.
The law is still developing. In most major patent jurisdictions, the current position is that inventors must be natural persons. But that does not answer the more difficult practical question: if a human inventor used an AI tool as part of the inventive process, how much AI involvement is too much? At present, there is limited guidance on where the line sits between using AI as a brainstorming aid and relying on AI to generate a material part of the inventive concept, including embodiments.
One way to address this may be to put guardrails around AI use now. In-house teams should consider updating invention disclosure forms to ask whether generative AI was used, which tool was used, whether any confidential information was entered, what prompts and outputs were relevant, and whether the tool contributed to the technical concept. They should also encourage inventors to keep records of relevant AI conversations and outputs, just as they would keep lab notebooks, design notes or invention records. And, at a minimum, organisations should make clear that public AI tools must not be used for confidential invention-related work.
2. Geopolitics is affecting patent strategy
Five years ago, many companies could still think about innovation, manufacturing and patenting in broadly global terms. Research could be done where the talent was. Manufacturing could be placed where it was most efficient. Patent filings could be directed to the major markets, the key enforcement jurisdictions and the countries where competitors were active. Recent geopolitics are beginning to strain that model.
The current geopolitical environment is pushing governments to build local capability in technologies they consider strategically important. Semiconductors are the clearest example. The US CHIPS and Science Act included about US$52 billion in funding for semiconductor manufacturing, research and supply chain resilience. The European Chips Act originally aimed to mobilise €43 billion by 2030, and the European Commission has more recently referred to more than €52 billion in committed public and private investment. Quantum technologies are another example. In June 2026, the White House issued executive orders on quantum innovation and post-quantum cryptography, including Executive Order 14413 on “Ushering in the Next Frontier of Quantum Innovation”, which directs a whole-of-government approach to accelerate quantum computing, sensing and networking, promote commercialisation and deployment, support domestic supply chains, and protect sensitive technologies. Similar policy attention is being directed toward critical minerals, clean energy, advanced manufacturing and defence.
At the same time, governments are tightening export controls, security obligations and technology controls around many of the same areas. The June 2026 White House quantum orders are useful examples because they do both things at once: one order seeks to accelerate US quantum capability, while the companion order on post-quantum cryptography treats quantum computing as a concrete security risk and sets federal migration requirements for cryptographic systems. Recent US controls on quantum technologies, advanced semiconductors, semiconductor manufacturing equipment and AI computing point in the same direction. The policy aim is not only to support local industry, but also to restrict access to technologies that governments see as relevant to national security, military capability or strategic competition.
This may be relevant for patent strategy in two ways. First, manufacturing locations may change as companies respond to subsidies, procurement preferences, supply chain resilience requirements and local capability policies. If the places where products are made, sold or improved change, filing strategies will need to change with them. A jurisdiction that was once commercially marginal may become more relevant if it becomes part of the manufacturing or supply chain story.
Second, tighter export controls and secrecy regimes may reduce patenting activity in some sensitive technology areas or at least change the timing and geography of filings. Some inventions that would previously have been patented may be retained as trade secrets. Others may require foreign filing licences, non-publication requests, secrecy orders or careful sequencing before any overseas filing occurs, if at all. In-house teams should not assume that the historical answer to file broadly and publish in the usual way will always be available (at least in some technology areas).
3. Shifts in technology ecosystems are affecting patent strategy
For many companies, international filing strategy has historically followed a familiar pattern: the US, Europe, Japan, Korea and perhaps a small number of additional markets such as Canada or Australia, depending on budget and commercial plans. That approach may still be right for some portfolios. But in-house IP leaders should be testing whether it still matches where the relevant technology is gaining momentum.
Patent filing activity in parts of Asia has been increasing steadily over the past 5 years, and the commercial reasons are not difficult to see. Supply chains are diversifying. Many businesses are looking beyond China-only manufacturing strategies. Digital adoption is high across the large younger populations in Asia. Patent offices in the region are also improving their systems and international cooperation.
India is the clearest example of the shift. Patent filings in India reportedly rose 30.2% in FY2025–26. Over the last five years, filings have increased by 146%. India is now described as the world’s sixth-largest patent filing jurisdiction, with domestic applicants accounting for about half of filings. This is a jurisdiction with increasing innovation density and rising strategic relevance.
This does not mean businesses should consider filing patent applications for all patent portfolios in all these Asian countries. Several South and Southeast Asian jurisdictions are also becoming more relevant for particular technology areas because of their manufacturing base, policy focus, investment activity or commercial adoption. Singapore may be relevant for semiconductors, AI, data infrastructure, fintech, telecoms and other high-value digital technologies, particularly where applicants want a sophisticated IP system and a regional hub. Malaysia is increasingly relevant for advanced packaging, semiconductor manufacturing, EV components and battery-adjacent technologies. Thailand is compelling for automotive, EVs, batteries and charging infrastructure, given its established automotive base and push to localise EV supply chains. Indonesia is important to consider for battery technologies, nickel processing, cathode materials, energy storage and EV supply chain innovations, although enforcement, regulatory and market-specific considerations should be assessed carefully. The Philippines may be relevant for semiconductor assembly, testing and packaging, electronics manufacturing, smart manufacturing, software-enabled manufacturing systems and digital services. Vietnam is also increasingly relevant for electronics manufacturing, telecommunications, embedded software, semiconductor assembly and testing, IC design and high-tech manufacturing. India sits slightly outside the Southeast Asia bucket, but for many applicants it belongs in the same strategic conversation. It is increasingly relevant for software, AI, telecommunications, electronics, semiconductors, medtech, pharmaceuticals, clean energy and advanced manufacturing.
The key point is not that every portfolio needs protection in all of these jurisdictions. It is that filing strategies should be tested against where the relevant technology ecosystems are developing, including manufacturing capability, supply chain movement, customer adoption, competitor activity, government policy and realistic enforcement prospects. A jurisdiction that looked optional five years ago may be strategically important now if the technology, market or supply chain has moved there.
4. The UP and UPC systems are becoming real strategic options
Three years in, the Unitary Patent (UP) and Unified Patent Court (UPC) are now live systems that patent owners should factor into European filing, validation and enforcement strategies.
The statistics are still developing, but they already show real use. Uptake of the Unitary Patent has been gradual rather than explosive. EPO data shows that the uptake rate for granted European patents converted into Unitary Patents was close to 30% in 2025, with uptake by European applicants closer to 40%. The top patentees using the UP include Samsung, Siemens, Qualcomm, J&J, and Becton Dickinson. However, the system is not only being used only by the largest multinationals. Almost one third of requests have come from SMEs, and around 8% from universities. The strongest technology fields include medical technology, computer technology, digital communication, transport and measurement.
The UPC side is equally important. One recent set of UPC data recorded more than 640 first instance actions after 32 months of operation, including more than 480 infringement actions, more than 80 revocation actions and more than 80 applications for provisional measures. These numbers are not yet mature enough to support sweeping conclusions, but they are enough to show that the UPC is no longer just an experiment.
It is interesting to note that early fears of mass central revocations have not quite played out in the way some expected. Revocation-only cases remain a minority of UPC activity, while infringement actions make up the larger share of the docket. The courts are also moving quickly, with many infringement actions reaching oral hearing within about 15 months, and preliminary injunction proceedings capable of being decided in roughly 10 to 16 weeks. That speed can be attractive for patentees seeking efficient enforcement, but it also increases the pressure on defendants and requires careful litigation planning.
For patent owners, the point is not that UP protection is always good or always bad. The top 5 patentees using this system are not automatically converting all of their granted EP patents into UPs, but only a portion of them, suggesting that Europe now requires more deliberate portfolio-level decisions. Which patents should be opted out? Which might benefit from central enforcement? Which are too commercially important to expose to central attack? These questions should be answered while there is still time to make considered choices.
What should IP leaders do now?
- Update invention disclosure forms to capture AI use, tool identity, prompts, outputs and confidentiality issues.
- Consider whether government policy, export controls, secrecy rules or funding conditions should affect patent-versus-trade-secret decisions and the sequencing of international filings.
- Re-test filing jurisdictions against future manufacturing, supply chain, enforcement and market relevance, not just historical habit.
- Make deliberate UP and UPC decisions at portfolio level before a dispute arises.
