Subscribe: Email | TuneIn | RSS | More
My co-host Ken Suzan and I are welcoming you to episode 179 of our podcast IP Fridays! Today’s Interview guest is Peter Berg. If you don’t know Peter, he is the Chief IP Counsel at Infineon Technologies and is also serving as the VPP president. We talk about the industry association of in-house practitioners VPP, politics, patent strategy at Infineon, litigation ready patents, NPEs, acquisitions, the German emloyees invention act and many more things. So stay with us for this very interesting interview.
Before we jump into this interview, I have news for you!
The USPTO has issued its first AI-related disciplinary action, published on August 27, 2026 in In re Brian E. Mitchell (Proceeding No. D2026-16) before the Office of Enrollment and Discipline. The underlying settlement was signed by the attorney on July 20, 2026 and by the OED Director on July 21, 2026, with final approval given on July 27, 2026 by Tricia Choe on behalf of USPTO Director John A. Squires. The action resulted in a public reprimand against California patent attorney Brian E. Mitchell after a generative AI tool produced a claim comparison chart containing fabricated citations to the patent’s internal file history. The case underscores that AI-generated hallucinations must be verified by a human before filing.
InterDigital has escalated its patent dispute with Disney across three fronts within a month. On August 25, 2026, InterDigital filed two new actions at the Unified Patent Court: a preliminary injunction request at the Mannheim Local Division (UPC_CFI_3044/2026) and a main infringement action at the Düsseldorf Local Division (UPC_CFI_3043/2026), covering Germany, France and the Netherlands. Both target Disney+ Premium’s switch to the VP9 video codec, which InterDigital says still infringes patent EP 2 465 265, a patent the Mannheim division had already found valid and infringed in an earlier ruling that also branded Disney an unwilling licensee; the Düsseldorf action adds a second patent, EP 2 605 473, covering the Up Next continue-watching feature. On September 2, 2026, the Düsseldorf Local Division granted InterDigital an injunction against Disney covering Germany and the Netherlands over a separate patent protecting seamless playback continuation when switching between devices, upholding the patent’s validity in the same ruling. It was InterDigital’s third UPC injunction against Disney. Then, on September 15, 2026, InterDigital filed a damages claim for EUR 101.7 million against Disney at the Munich Regional Court I (case no. 7 O 8870/26), based on a November 2025 finding that Disney+ infringes InterDigital’s HDR video patent EP 2 132 923; the claim also covers viewers in 19 European countries served through a Frankfurt-based server location. For companies relying on codec or connectivity technology, the dispute is a reminder that switching to a different standard does not automatically avoid infringement if the asserted patent is drafted broadly enough.
The EPO’s Enlarged Board of Appeal ruled on September 3, 2026 in G 1/25 that a patent description only needs to be amended to match amended claims if a discrepancy between the two causes a requirement of the EPC, such as clarity, novelty or inventive step, not to be met.
The EPO’s Board of Appeal 3.3.08 fully revoked Moderna’s key mRNA patent EP 3 590 949 following a two-day hearing that concluded in mid-September 2026, on the grounds that the claimed subject matter extended beyond the original application. The revocation removes the basis for Moderna’s earlier wins against BioNTech and Pfizer in Düsseldorf and London.
On September 15, 2026, US sportswear maker New Balance filed a trademark suit against Decathlon’s US units, Decathlon America LLC and Decathlon USA LLC, in the US District Court for the District of Massachusetts (case no. 1:26-cv-14235). New Balance argues that the K logo on its Kiprun running shoe line, when mirrored, reads unmistakably as an N and imitates the N mark it has used for roughly 50 years, claiming post-sale confusion.
On September 8, 2026, the Court of Justice of the EU ruled in Case C-298/23, Inter IKEA Systems v Vrijheidsfonds, on a referral from the Brussels Enterprise Court, that the owner of a well-known trademark can prohibit its use by a political party unless the party sets out the specific reasons why its freedom of expression outweighs the trademark owner’s interests. The case arose from the Vlaams Belang party’s 2022 campaign promoting a migration plan styled as IKEA assembly instructions.
And now for the interview with Peter Berg!
In this episode of IP Fridays I talk with Peter Berg. He is Chief IP Counsel at Infineon Technologies and President of VPP, the German association for in house IP professionals, since May 2025. Berg is an electrical engineer by training and leads a team at Infineon that manages a portfolio of more than 12,000 patent families, while the company simultaneously enforces its rights offensively in Germany, the US and China. That combination of hands on portfolio work and association leadership makes him one of the rare guests who can speak to both the operational detail of one of Europe’s largest patent portfolios and the political side of Germany’s IP landscape.
The guest: Peter Berg and the VPP
Peter Berg is an electrical engineer who became Chief IP Counsel at Infineon Technologies and, since May 2025, President of the VPP, the association representing in house patent professionals in Germany. His team manages a portfolio of more than 12,000 patent families while running an offensive enforcement program across Germany, the US and China. That mix of operational and political experience is exactly why I wanted him on the podcast.
Germany’s national IP strategy: more than a promise on paper
I asked Berg what has become of the national IP strategy promised in Germany’s coalition agreement, and what happens if it does not materialize during this legislative term. His answer was clear. The question is no longer whether the strategy will arrive, only how comprehensive it turns out to be. If it falls short of what industry needs, the profession will simply push for adjustments afterward.
Berg names three levels that need attention. First, society as a whole, from kindergarten to retirement, needs to understand that intellectual property underpins our prosperity rather than being a line item in the accounts. Second, academia and professional education. Third, startups and the broader Mittelstand, where well funded structures are still missing. The German Patent and Trademark Office has an educational mandate but far too small a budget for it, and the regional patent information centers are underfunded as well. Berg makes a point that is easy to miss from outside Germany: the patent office reports to the Ministry of Justice rather than to an innovation focused ministry, even though its real purpose is to strengthen the economy. That three ministries, research, economy and justice, are now working on the strategy together strikes him as genuine progress.
Litigation ready patents: how Infineon manages 12,000 patent families
Infineon has roughly tripled its turnover since 2017 while keeping portfolio size largely stable. The answer, Berg says, is quality over quantity. Inventors pitch their ideas to a panel of senior technical experts and patent attorneys, and the process does not stop there. The draft itself is reviewed by multiple stakeholders, and once the notice of allowance arrives and the patent is granted, it goes through another review. The criteria stay consistent throughout: technology assessment, market size, detectability, circumventability and fit with Infineon’s own roadmap.
Internal discipline alone does not make a patent litigation ready, though. That requires real experience with conflict, and most of that experience comes from the defensive side, where Infineon has plenty of practice finding weaknesses in patents asserted against it. Being on the offensive, Berg says, is a different level of learning entirely, whether in court or in licensing negotiations where the other side is not sure it would get a fair outcome if the case went to trial.
In practice this also means systematic pruning. Internally Infineon calls the relevant phases US8 and US12, the years when annuity costs peak in the US portfolio, by far the most expensive part of the estate. Berg says his team knows every one of the more than 12,000 patent families in detail, and offensive enforcement is focused on the markets that matter most for electronic products: Germany, the US and China.
An injunction at a trade show: national courts versus the UPC
One concrete trigger for this conversation was an ex parte preliminary injunction Infineon obtained at a German trade show, without the defendant having filed a protective letter in advance. For Berg, enforcement at trade fairs is nothing new, it has been part of the playbook for years, now extended to venues outside Europe that Western companies have used less often in the past.
His view on the UPC is nuanced. He credits the court with a strong start, good management and judges selected for quality rather than politics, and praises the body of case law it has built in a little over three years. On the other hand, some member states are still missing, a comprehensive body of case law takes time to develop, and costs do not fit every business case, especially where achievable damages barely justify the expense. Infineon in any case validates most European patents in only a handful of countries rather than dozens, so the practical difference between national and UPC enforcement is smaller than it might seem. For smaller cases the option to opt out and litigate nationally alongside UPC rights remains important, even though systematically double patenting is not something Infineon pursues for cost reasons.
Four systems at once: coordinating cross border enforcement
Infineon’s legal and patent departments have grown together over 25 years into a fully integrated international organization, with teams in Europe, North America and several Asian countries that, by Berg’s account, communicate on close to a weekly basis. Add to that a strong network of outside counsel. The commitment to management is not to shy away from any jurisdiction where the need arises.
Berg is careful that patent attorneys do not spend their entire careers doing only litigation or only prosecution. Attorneys who rotate into a litigation team bring that experience back into their regular business support afterward, and that cycle in turn improves the quality of new filings. Litigation has become part of daily business, but never something a team slides into unprepared. Going on the offensive against a competitor requires planning several steps ahead.
IP due diligence in practice: the Marvell Brightlane deal
I found Berg’s answer on the Marvell Brightlane acquisition, Infineon’s 2.5 billion dollar deal for Marvell’s automotive Ethernet business that closed in August 2025, particularly revealing. In the semiconductor industry, he explains, IP is rarely primarily the patent portfolio. What matters more are frozen technologies and so called IP blocks, reusable design blocks that enable fast product development. Infineon almost never acquires a company because of its patents. Being a large, established player, its own portfolio is usually broad enough to secure a new technical area against incumbent competitors.
The real work starts after signing. It means listening without being pushy, understanding the new colleagues’ processes and tools before deciding how to integrate the portfolio. Since none of Marvell’s IP staff joined Infineon in this deal, the team had to build up portfolio expertise using the acquired engineers and set up an operational patent committee, then plan which areas of the portfolio were already strong and where investment was still needed. The faster that integration decision gets made, Berg says, the better, because product development can never be allowed to stall because of an acquisition.
The Cypress lesson: NPE exposure from legacy portfolios
One risk many companies underestimate in acquisitions is patents an acquired business sold to a non practicing entity, or NPE, before the deal closed. That happened with Cypress Semiconductor in 2016, after its merger with Spansion left it with an oversized portfolio it did not need. Cypress sold around 2,000 patents to the NPE organization IP Value, and only a few years later Infineon acquired Cypress. Because the sale included a retained license for Cypress’s own business, Berg sees no exposure for Infineon from those specific patents today.
The broader issue of NPE enforcement, though, remains real. Realistically there are many dozens of new cases a month, often directed at customers or the wider ecosystem rather than at chipmakers directly. Even when Infineon is not the target, it has to support its customers whenever its own technology is implicated. For the industry as a whole this consumes a great deal of resources without moving innovation forward, a point that comes up especially often at this time of year in Munich, when IP practitioners from around the world gather for meetings.
Post quantum cryptography: patent density ahead of standardization
In December 2024 Infineon became the first company to achieve a CC EAL6 certification in post quantum cryptography, one of the highest security certifications available. Berg describes two angles to the work. On one side, Infineon develops its own chipsets to enable quantum computing itself, already a real business rather than pure research. On the other side, customers across Infineon’s application fields need to be prepared for what he calls quantum day, the point at which powerful quantum computers become available, including to bad actors. Because that day may only be three to five years away, infrastructure needs to be built now so it can handle new cryptographic approaches.
On standard essentiality he is measured rather than evasive. There are certainly patents in this field from Infineon and others, but security standards have historically leaned toward royalty free licensing rather than the large patent clusters seen in cellular communication from 2G through 5G. Whether that pattern repeats for post quantum cryptography, he cannot say, but Infineon is preparing for either outcome.
Germany’s Employee Invention Act: opportunity or burden
Asked whether Germany’s Employee Invention Act is an opportunity or a burden for the country’s innovation ecosystem, Berg does not hold back. With more than 30 years of experience, he estimates that 99 percent of all inventions are simply the regular work product of engineers in R&D, requiring no extra incentive at all. In countries without an equivalent law, employees are demonstrably just as innovative and just as willing to share their inventions with the company.
His core objection is that invention does not equal innovation. Procurement, production, quality management and many other functions all contribute to a product’s success, not just the named inventors. That the inventor alone receives extra compensation while everyone else does not strikes Berg as unfair, especially since Infineon already rewards its top engineers appropriately regardless of statutory requirements. He calls the law a bureaucratic nightmare that should be abolished or at least radically simplified. Smaller organizations, he notes, often just ignore it or implement it badly, and every company Infineon has acquired with German operations has come with its own legacy mess to clean up.
Artificial intelligence in day to day patent work at Infineon
On AI, Berg describes three layers. The first is established tools: patent search, translation between the team’s main working languages, and, for about two years now, drafting replies to office actions. As with any tool, the benefit only shows up with repeated use at real volume, which is why Infineon works deliberately with outside counsel to broaden adoption across the team.
The second layer is ongoing testing of new tools, limited mainly by the time available to evaluate them. The third is everyday office use. Since spring 2025, everyone from paralegals to attorneys has been able to use AI tools broadly within a secure environment. The bar stays the same throughout: if AI output is not at least as good as an attorney’s prior work, it is not worth using, or the cleanup afterward eats up any time saved. Where it does work, responses to standardization proposals, for instance, can be produced far faster. One question Berg expects to face next year, without yet having an answer: how should efficiency gains be split between the company and outside counsel once both sides benefit from AI?
Three messages for the IP community
I closed by asking Berg for the most important messages he would want to give our listeners, mostly IP professionals in companies and law firms. His first message: think about IP strategy from the end, starting with what needs to be asserted against whom and for what purpose in ten or twenty years, rather than letting today’s deadlines dictate everything. Only that kind of multi year plan, combined with a sound cost benefit ratio, produces genuinely litigation ready patents.
His second message is again about artificial intelligence: anyone who has not started training with it yet risks missing the wave, and this is not a skill built overnight. His third message is aimed at anyone not yet connected to peers in the profession: use the associations and networks that already exist to build that exchange, a thought that ties neatly back to the VPP, where our conversation began.
And here is the full transcript:
Peter Berg: Hi, I’m Peter Berg, Chief IP Counsel at Infineon Technologies and President of VPP, the Association of In-house Patent Practitioners in Germany. And you are listening to IP Fridays.
Rolf Claessen: Today’s interview guest is Peter Berg.
If you don’t know Peter, he is an electrical engineer by training and is the Chief IP Counsel at Infineon Technologies and also serving as the VPP president. Thank you very much for being on IP Fridays.
Peter Berg: My pleasure.
Rolf Claessen: So first I want to talk about VPP. You succeeded Beat Weibel from Siemens as VPP president in May 2024 and have publicly called for Germany to finally deliver on the coalition agreement promise of the national IP strategy. So politics, what are the two or three most concrete things that strategy must contain? What happens if it doesn’t get done in this legislative period?
Peter Berg: First of all, I think I need to correct you. It was May 2025, so not that long as a standing president. But if you mention national IP strategy, national means here for Germany. I think the question isn’t anymore whether or what happens if we don’t get that strategy. I think that is now on its way. The question is rather how good and comprehensive the national IP strategy will be, and if it’s just not enough from an industry perspective, I would say then we adjust afterwards. Of course, there is a big paper out from the BDI, the German Industry Association, published in 2024, and I need to focus on something which is really close to my heart. If you ask me the main content it should have, it is first of all better IP education, and I see there are kind of three levels. One is really society as such, so each and everybody from kindergarten to retirees, we need to acknowledge in the knowledge society that IP is central for our well-being. We have to respect the IP of others. We have to be proud to create IP, not just treat it as an entry in the accounts. Yeah, society at large. The second level is really academia and education in our professional environment. And finally, life is difficult for startups, people who want to create a new company, and small and medium-sized enterprises. There is good stuff out there. The German Patent and Trademark Office has a mandate to educate, but a very small budget to do so. In a federal country like Germany, there are local support systems, be it the chambers of commerce or the patent information centers. The structure is there, but it’s underfinanced, and Germany has to gear up on that one. So besides IP education, I believe solid national IP institutions are key. This is also a funding issue, be it headcount or technology. Think of the German Patent and Trademark Office, think of the courts, be it on the infringement side or the nullity side, so the Federal Patent Court. We need to back up nationally the European institutions, since we believe not every case is worth taking to the UPC, so for smaller cases we need a good national system as a backup. And government needs to have that in mind and not rely too much on European institutions alone. Finally, facilitate the IP transfer from universities and research institutions into industry. Right now it’s all about startups, et cetera, but that kind of IP transfer starts first of all with people leaving university, leaving those research institutions and joining industry, be it as staff or through research contracts. And finally, of course, startups are the area where Germany needs to catch up in particular. Perhaps I need to add for non-German listeners, there is a particularity in Germany in that the IP institutions, like the Patent and Trademark Office, report into the Ministry of Justice, although it’s an instrument for thriving innovation and building a solid economy. So it’s right now a totally new thing for this country that three ministries, research, economy and justice, are working together on this national IP strategy, with only recently some good progress.
Rolf Claessen: You’ve said publicly that Infineon’s primary focus is patents that provide clear competitive advantages, and that for you it’s about litigation-ready patents. That’s a very specific phrase. How do you define it operationally, and how has the pruning of your more than 12,000-family portfolio changed in the last two years?
Peter Berg: Yeah, I think that’s a process which lasted much more than two years, and it’s never over. This is work in progress. Thinking back to 2017, I joined Infineon, and the company has developed very nicely since. I think we roughly tripled turnover but kept the portfolio size roughly stable, despite acquisitions, despite that growth, also in R&D, et cetera. We’ve even been able to shrink costs since then, and I believe we’ve achieved our strategic goals, now with a good, or let’s say not yet good enough, cost-benefit ratio. And the answer to that equation is quality. Now, we as patent practitioners might call it differently, but at the end it comes down to the fact that, most likely for all companies and for practitioners who support companies in private practice, you need a collaboration platform. That means for Infineon, the inventors are key. They are involved in decision-making, so they pitch their invention to senior technical experts, my team members, and patent attorneys, and the quality of that expert panel makes a big difference. Then come repetitive reviews, not just once the decision to file is taken, but the draft itself is also reviewed by multiple stakeholders. Then we get the notice of allowance, the patent is granted, another review, and later on we try to apply consistent criteria: what’s the technology assessment, what’s the market size, is it detectable, is it easy to circumvent, et cetera, and does it fit with our own roadmap. And of course it would be nice if you could easily compare it with a competitor’s roadmap, or take the time to speculate on certain market changes which might be disruptive. And I believe if you apply that consistently over a sufficiently long period, you get patents you could call litigation-grade or litigation-ready. What is then missing, though, is real experience with conflicts against competitors and how the courts look at that. So the truth comes out once you go to court. On the defensive side, of course, we have tons of experience trying to find weaknesses in patents being asserted against Infineon, et cetera. But I think it’s another level of learning to be on the offensive, and that might play out in court, but it might also play out in licensing negotiations, where the other side is sufficiently unsure whether they would get an outcome they believe would be fair if they went to court with those patents. You asked me about the pruning of all this. So what is our approach there? There is a kind of forced pruning rate which we believe is necessary, to make it mandatory for everybody to identify the weakest patents in the respective portfolios. We focus on the years where costs are highest, what we internally call US8 and US12, since a big chunk of money needs to be spent on annuity payments for the US portfolio, which is the dominant one, and for the rest of the world it’s rather towards the end of the patents’ lifecycle. Now, having a good team, people who really like the technology and stay in charge for a long time, I could proudly say that my team knows every one of those 12,000 patent families well, and not just on the German or the US side, but really focuses on the jurisdictions with the biggest markets for electronic products. And there we see offensive litigation as a good choice, and that is Germany, the US and China.
Rolf Claessen: Yes, talking about patent quality is really a hot topic right now, especially in industry and at the EPO, I guess, but that’s a topic for a different hour, probably. But talking about litigation-ready patents, you recently obtained possibly the first ex parte preliminary injunction at a German trade show in recent history, without a protective writ having been filed by the defendant. What does that show about your view on German national courts versus the UPC for this kind of enforcement?
Peter Berg: I mean, first of all, litigation is not an end in itself, and in the markets a semiconductor company like Infineon operates in, a product might implement potentially hundreds or thousands of patents. So litigation itself doesn’t resolve all of that, but it has its place in our strategy execution. You’ll understand that I can’t get into specifics of any of the litigations you hinted at. But enforcement of trademarks, and I think you mentioned a patent case, trademarks and patents at trade fairs has been part of our activities for years, and sometimes you can catch somebody by surprise. It’s all about good preparation. But what I believe is more important is that we also prepare for venues so far less used by Western companies, be it China, et cetera. So being litigation-ready means looking into different venues. The UPC is certainly one of them, but there’s more beyond that. Enforcement for us is not only a German or European measure. I personally believe national courts and the UPC will both have a role to play in the future. There are pros and cons to each, which need to be evaluated case by case. I need to congratulate the UPC, which had a great start, good management, good communication about what they do, and a selection of judges based on quality, not politics. In a little more than three years, they’ve built up a pretty good body of case law. On the downside, some countries are still missing, building a really comprehensive body of case law will take time, and the cost side is still a concern, meaning that not every kind of business case reaches damages high enough to justify it, and there’s a certain risk that the case won’t be decided quickly by the court. In such situations, it might not make sense to go to the UPC. We have, I guess, quite a few companies in the electronics or mechanical engineering field where we don’t validate European patents in dozens of countries, maybe two, three, four, and over time it might be even fewer. So geographically, the difference between US, UPC and national enforcement isn’t that big. That’s why the alternative, and I mentioned this a bit at the beginning regarding the national IP strategy, needs to stay a good alternative. So the possibility to opt out and litigate nationally in lesser business cases remains, from my perspective, important, and we continue to use that as well.
Rolf Claessen: Yes, and also the hurdle for double patenting has been lowered a little, so you can still have national patent rights in parallel with UPC patent rights. That’s also an advantage, probably.
Peter Berg: Yeah, I think that was almost an invitation to say, you know, if you’re not sure whether the unitary patent is your choice, you can do both in parallel. True. But again, all we need to do is achieve our goals with a good cost-benefit ratio, so systematically double patenting is not really the way forward for us.
Rolf Claessen: Right. Yeah, that was a very interesting insight about German national courts versus the UPC. But in fact you’re also litigating across borders, in many different countries, for example the US, Europe, Germany, and other countries, in China maybe. So what does coordinating enforcement across all these different systems simultaneously actually look like from the inside, and was it always the plan to have such complex litigation and enforcement systems, or did it evolve? What’s your take on this?
Peter Berg: Yeah, at Infineon, our legal and patent departments, as we call them, have grown together over the last 25 years, roughly the age of our company, into an organization that doesn’t just have broad general legal or semiconductor application market expertise, for us that’s automotive, consumer goods, industrial and IoT, but is an organization that’s been tested by hundreds of assertions against us and our customers, and by negotiations with operating companies and NPEs. So that comes with experience. If I look now at our organization, not just my own team but legal in general, we have patent and legal teams in all the hotspots, and that’s grown in the last couple of years too. It’s Europe, of course North America, and various countries in Asia, and it’s a fully integrated international organization. We know each other and communicate, it feels like, on at least a weekly basis. Not to forget a good network of outside counsel to execute enforcement wherever strategic needs require it. That’s kind of our promise to management: we don’t shy away from any jurisdiction, if the need is there we can act and set up a team. Perhaps we’ve learned more on the defensive side, which we can now translate into offensive action. And I guess you’re somehow relating your question to a specific litigation against a Chinese competitor, and that’s not a litigation you kind of slide into. You need to plan a few steps ahead. Litigation has become part of our daily life, and the key moment is always, once it’s decided that this is going to happen, to set up a team for the new litigation. It starts to feel like a routine operation. The staffing depends of course on the technical area of the enforcement, the jurisdiction, and the workload of existing attorneys. And the good thing is that once we rotate patent attorneys into a litigation team, it’s great experience for them, that leads to expertise, and since we try not to over-specialize, so that people don’t only do litigation or only do prosecution, that experience from litigation is a great asset when they go back to supporting a certain business unit doing prosecution work. And coming back to your first question, that personal experience also helps in creating litigation-ready patents over time.
Rolf Claessen: Yes, you’ve also acquired a couple of companies, and I have a question about a certain deal, especially the Marvell Brightlane acquisition and the IP due diligence. This $2.5 billion Brightlane deal closed in August 2025 and presumably transferred a substantial patent portfolio in the automotive, let’s say networking, technology field. How much of that big deal is really about buying IP, and how does a deal like that change your daily work in your IP function the day after closing? So once you’d signed the deal, did your daily life change?
Peter Berg: It depends on the size of the deal. A couple of questions in there. Let me start with IP in the semiconductor industry, my industry, and that’s more than patents, trademarks and design rights. Speaking to practitioners and R&D colleagues, IP for them is rather frozen technologies and what we call IP blocks, design blocks which enable fast product designs that you can reuse. And if I take that as a basis, although each acquisition is different, IP plays a very big role because it should enable us to bring new products to market quickly, et cetera. Seldom does Infineon acquire a business because of its patent portfolio. Being a large, established player, our existing patent portfolio should be good enough to move into whatever technical area is of interest, and to make sure incumbent competitors couldn’t easily attack us when entering that field. Now, you mentioned the Marvell acquisition. We call it rather the Marvell automotive Ethernet acquisition. From Marvell’s perspective it was a smaller part of their business, which they sold to Infineon about a year ago. How should I say it, you get the assets but you get relatively little IP. So it’s more that you try to familiarize yourself with the assets and how they fit our business case, et cetera. Of course there is due diligence to get an idea and to make sure that everything that should belong to the business is transparent, but the more important part comes after the acquisition. Once the business is owned by Infineon, all the other functions handle their own areas of responsibility. For us, when it comes to IP rights, it’s about reaching out to the new colleagues, helping them for the first couple of days, but also leaving them alone, because there are so many changes they have to go through, and not being overly pushy. Understanding their current processes and tools, and their view of their own portfolio, is important at that point. So you hold back a little and see what they do and why they do it, sometimes you can learn from their best practices. But there comes a moment when we need to finally decide how to integrate the portfolio, and which of our processes and tools they have to adopt, and the faster you make that decision, the better. Since in the case of Marvell none of the IP staff joined us, the task was to build up a portfolio using the expertise of the senior engineers, get a good operational patent committee running, and develop a plan for which areas of the portfolio are good enough and which need to grow, in line with the growth of the business. It’s now more than a year ago, but that was an interesting time, and I doubt many people listening are aware of it, but a big thank you to everybody involved back in 2025.
Rolf Claessen: From your answer I get that your job involved a lot of listening, basically, to the needs of the teams that joined your company, right?
Peter Berg: At least as much talking as listening. Of course they want to know as well how we do things and what’s expected of them. So first of all it’s about reaching those people and making sure, and this is also a concern, that if everything changes for them, then perhaps the patenting process isn’t the most important thing. Product development is key, and making sure there’s no interruption so they can keep their timelines, et cetera. So being very mindful of that setting. But you’re right, it’s not just about telling them, it’s about listening to them and finding the right moment to switch.
Rolf Claessen: Right, I have a very different question now, about NPEs, non-practicing entities, and exposure. I think some NPEs acquired patent families that trace back to Cypress, just before you bought it, or around the time of your acquisition. Are these patents, held by these NPEs, now a problem for Infineon?
Peter Berg: That would be it, right? Actually, no, let me give some context first. Companies in distress, or with an overly large patent portfolio for their business, sometimes sell patents at a certain point in their history. Regarding Cypress Semiconductor, that happened around 2016, after they merged with a company called Spansion and ended up with a large combined portfolio they didn’t need. At that point Cypress teamed up with an NPE organization called IP Value and sold them around 2,000 patents, that’s public knowledge, and only a few years later Infineon acquired Cypress. So that was done by Cypress before it became an Infineon company. But good practice is, if you sell a patent, you retain a license for your own business, so in that sense I don’t feel any exposure from those patents. But of course this touches on the general topic of exposure to NPE enforcement, which is a concern in the electronics industry. Realistically, hundreds is perhaps too much, but many dozens of new cases a month, often against our customers or the wider ecosystem. But since it involves technology implemented in our products, even if the enforcement doesn’t target semiconductor players directly, we need to support our customers, and that is a kind of a load we need to help carry. But these are typically patents that somebody else sold to an NPE. You were also asking whether acquiring businesses brings new opportunities as well as new risk. I think that’s natural, and we shouldn’t shy away from that kind of risk. If we, like any other company, go for the big, high-margin, fast-growing businesses, the stakes are high in those areas, and you can’t complain if these also turn out to be litigious marketplaces, because many people want to be there. Around this time of year, since it’s Oktoberfest season here in Munich, many IP practitioners from around the world gather for meetings and share their view on the world. Everybody talks about AI, and NPEs try to acquire whatever patents are on the market that could be the basis of stronger or weaker infringement actions against that part of the industry. From an industry perspective, this consumes a lot of resources and doesn’t really help accelerate innovation in our space.
Rolf Claessen: Good that you have a solution for the patent families that were acquired before you acquired Cypress. And now, turning to a very different topic, cryptography. You’re also working on quantum cryptography and post-quantum cryptography, a difficult word, and you, Infineon, were the first to obtain a certification, a really high certification in post-quantum cryptography, the so-called CC EAL6 certification, a very high certification in cryptography for our listeners, in December 2024. Now this field of technology is heading towards standardization. How patent-dense is Infineon’s patent stack in this field, and how are you thinking about the potential standard-essentiality of this new post-quantum cryptography technology, where maybe the American standards from the NIST institute become the infrastructure, or the standard?
Peter Berg: It’s a big field with different components. We take the quantum challenge seriously at Infineon, but there are two elements to it. You spoke more about cryptography, but for us, as a semiconductor player, on one hand we want to enable quantum computing with our own chipsets, and we support two different technology streams there, it’s not just development, it’s already a real business. On the other hand, we need to educate and enable customers in the application fields we serve to prepare for that quantum day, the day quantum computing becomes available, and not only to the good guys. You need to think about this now, because it might be just three, four, five years away, you don’t know for sure. It means investing in infrastructure able to handle post-quantum cryptography and security in communications, because you need certain computing power and different ways of handling cryptography algorithms compared to today. Today’s approaches won’t be enough, you need better ones, and you mentioned one at the beginning, which Infineon has proposed. That’s not easy. If you invest in a core today, you’re investing in something like a power station, which stays connected to the internet for a long time, so you need infrastructure robust enough to handle post-quantum cryptography. Regarding the patent stack question, yes, certainly there are several patents in this field owned by Infineon, but also by others. In general, I believe, both in semiconductors and in security more broadly, this tends toward royalty-free licensing, and standard setting has in the past been marked by the avoidance of the big patent clusters we saw building up in cellular communication, from 2G through 5G. I don’t see that happening in the post-quantum cryptography field, but who knows, so we prepare for whatever scenario comes.
Rolf Claessen: Let’s see if the Infineon patents end up in a big pool or something later on, and you get all these license fees.
Peter Berg: I wouldn’t mind, yeah.
Rolf Claessen: Let’s see. Right, now turning to another topic, maybe, for example US colleagues ask me a lot about the German employee invention law. It’s very different from other countries, especially the US. Do you see the German employee invention law rather as an opportunity or a burden for Germany’s innovation ecosystem?
Peter Berg: As IP professionals in larger corporations in Germany, we’re so used to that law that we hardly challenge its usefulness. To your question, though, with more than 30 years of experience, I’d say that 99% of all inventions are made by engineers in R&D as part of their regular job, nothing extra on top, so they don’t need an extra motivation for it. And we see it in other countries too, where that kind of law doesn’t exist, and people are just as innovative and just as ready to share their inventions with us as an IP organization, with or without the law. So it doesn’t seem to fulfill the role some see as a competitive edge for Germany, and because of that inventor law, people just end up being more cautious, or more descriptive, about their work results. There are more reasons why I believe this is obsolete. Invention does not equal innovation. There are many other colleagues in a company, in procurement, production, quality management, et cetera, and all of them contribute to a product becoming a bestseller. It’s not a single person, and it’s not just one team’s result, and somehow it doesn’t feel fair that the inventor gets an additional bonus and the others don’t. In any case we’d give our top engineers a premium anyway, and we reward inventors in all other countries, with the law or without it, without any worldwide mandate for that kind of remuneration. So from a corporate perspective we don’t need a law to help us work out the benefits and share them with our top inventors. The law is a bureaucratic nightmare. It needs to be abandoned, or at least radically simplified. In smaller organizations, I guess it’s often just ignored or badly implemented. Believe me, every company we’ve acquired with operations in Germany has left us a bit of a heritage to clean up.
Rolf Claessen: Yeah, I understand. Now for a different topic, AI, the use of AI. We as patent attorneys talk about this a lot, it’s a big topic, how to use AI, whether to use AI, and how to use AI in law firms. But can you tell us about the use of AI in the operations of your team?
Peter Berg: Recently we discussed this and agreed on kind of three layers to it. First, good use of established AI tools. AI, in that sense, isn’t something that started three years ago with ChatGPT. It began with search, patent search, then translation, at least between the major languages like German and English, which most of my team masters, and then, I’d say about two years ago, drafting and replies to office actions. So there are established tools, and as with every tool, you need to use it repeatedly, at a certain volume, to really benefit from it. We try to work with the best of these established tools, driving wider adoption within the team and partnering with our law firms, since a substantial part of the legal work is done by our outside counsel. So this is a bit of a joint exercise. Second, there are constantly new tools coming up, which we try to test. There are many, so the time we have available to look into them is the limiting factor. I’d like to test more, because there might be a new frontrunner coming up that we should definitely try ourselves. But last, and certainly not least, we try to implement AI into everyday operations, into the office environment itself. The company made a good decision in spring to allow wide adoption of AI-specific tools in a secure environment, and that has opened up a lot of opportunities for everybody, paralegals and attorneys alike, and now it’s about fostering an open exchange on what’s possible and what’s not. And finally, since Infineon, like most corporations, doesn’t do everything in-house, there’s the question of what the benefits are that we can plan around, so that we get better quality. If the quality of the AI results isn’t as good as the attorney’s results before, then we shouldn’t use it, or we’d spend so much time cleaning it up afterwards that nobody could tell the difference. Second, we can now do things that weren’t possible before, because they were too tedious or took too long, or we get faster turnaround times, so we can respond more quickly to standardization proposals, what used to take many days we can now do in a few days, while still trusting the quality of the output. And only then comes what’s often discussed as efficiency gains, and the question that will probably be asked next year is how we split the pie: if both we and outside counsel benefit from using AI, what do we do with those benefits, how much stays with outside counsel, and how much should also benefit my budget?
Rolf Claessen: But I clearly see the advantages of using AI, I’m using it myself, especially for cases like oppositions, where in earlier times you had to read through some 30 papers, patents and books, and now you can just ask the AI where you find a certain feature across those 30 documents. So it’s really useful in a lot of cases. Now my last question to you, and maybe the most important one, what do you think are the three most important messages you want to convey to our listeners, who are mostly IP professionals, both in-house and in law firms?
Peter Berg: Do you want me to repeat things I said before, or can I think freely, in general? Free thinking, of course. Okay. Many of the things we spoke about now were about what to do next, and how to execute certain things, how to prepare. I believe that if you think about IP strategy, you think about it from the end: what do I need to assert, against whom, for what purpose, before building a portfolio of IP rights. So it’s not just about the deadlines you have today and tomorrow, which you somehow have to get through, but about taking a step back and thinking, what should this be good for in 10 or 20 years’ time. If you do that, it helps you get to litigation-grade patents and litigation readiness. We’re all busy, and there are so many things we need to do today and tomorrow, so it’s easily forgotten to look a little further down the road. So I’d say, think about a multi-year plan to achieve your goals, and there are hardly any moments, unless perhaps you’re one of the big AI platforms with unlimited resources, where that isn’t true. So plan for multiple years, but with a good cost-benefit ratio in mind. And I can’t really avoid mentioning AI here. It’s having a major impact on our operations. We’ve talked about it, we’ve tested it, but ramping it up is a different matter, and I think next year will be decisive. That’s really the full value chain, from interacting with our inventors to get better quality ideas, all the way to claim charting and negotiating, the bread-and-butter activities of patent attorneys, drafting office action replies is of course part of that too. I’m really keen to see that happen, and I’m trying to keep my team ahead of the wave. And I can only encourage everybody in our profession, if you don’t yet have a contact, team up with friends or colleagues in the profession, use the various associations that exist in our profession to network and discuss. The worst thing is to miss the train, and it takes time, this isn’t something you become good at overnight, so it needs time to train, and it’s good to start now.
Rolf Claessen: Yes, that’s a very good point, and closing where we began, with the VPP, exchanging ideas among colleagues really is key, I think, important. Thank you very much for being on IP Fridays, on our podcast. That was really interesting, and I think our listeners can learn a lot from this interview. Thank you very much.
Peter Berg: Don’t mention it, it was a pleasure talking to you, and for me it was also the first time doing this kind of podcast. Thank you.
Rolf Claessen: Thank you