Thank you for your enquiry. We will be in touch with you as soon as possible.

Stephan: At thyssenkrupp Intellectual Property, our motivation to integrate AI into IP management processes was twofold: to increase efficiency and to elevate the strategic quality of our work. We saw significant potential in using AI not just as a tool for automation, but as an enabler for more informed decision-making.

From a strategic standpoint, our goal was to support the business units more effectively by accelerating innovation cycles, improving the quality of invention disclosures, and enabling more precise patent analytics. For instance, AI allows us to monitor technological trends and identify white spots in the IP landscape—areas with untapped potential—much more systematically. This directly feeds into our strategic IP planning and portfolio development.

Furthermore, the use of AI helps us shorten the time between the emergence of an idea and the filing of a patent. With tools that support structured interviews or assist in the formulation of invention disclosures, we can engage inventors more efficiently and protect innovations faster.

Ultimately, we see AI not as a replacement for human expertise, but as a powerful complement. It enriches our processes, supports our inventors, and strengthens our ability to align IP strategy with business objectives—especially in a diverse and global technology group like thyssenkrupp.

How has the integration of AI transformed your team’s approach to managing and monitoring the global IP landscape—both internally and in relation to competitors?

The integration of AI has fundamentally enhanced the way we observe and navigate the global IP landscape. Internally, it has enabled us to work more proactively and systematically by leveraging data-driven insights. For example, we now use AI to analyze large patent databases more quickly and accurately than before, helping us identify relevant prior art, technology trends, and innovation clusters across industries.

This capability is particularly important in a diversified company like thyssenkrupp, where different business units pursue innovation in very distinct technological domains. AI helps us maintain a clear and up-to-date understanding of each unit’s IP positioning and detect opportunities for synergy or portfolio expansion.

Externally, AI tools have strengthened our ability to monitor competitor activities. We can now track new filings, detect emerging areas of patenting activity, and even identify potential IP infringements early—using technologies like image recognition or automated text analysis. This gives us a much stronger basis for strategic decision-making and risk assessment.

Can you describe how AI tools are reshaping the role of IP professionals at thyssenkrupp? Are new skill sets becoming essential for future IP leadership?

AI is definitely reshaping the role of IP professionals at thyssenkrupp—from traditional gatekeepers of legal protection to strategic enablers of innovation. As AI systems take over more repetitive or data-heavy tasks, such as prior art searches or trend analyses, our teams can focus more on the high-value activities: identifying innovation potential, advising on strategic IP positioning, and engaging more closely with inventors and business leaders.

This shift naturally brings with it a new set of required competencies. Future IP leadership needs to be fluent not only in legal frameworks but also in data literacy and digital tools. Understanding how AI models work—what they can and cannot do—is increasingly important. It's not about coding, but about being able to critically assess AI outputs and integrate them into strategic decision-making.

Interdisciplinary thinking is also becoming essential. Many of our most impactful AI use cases lie at the intersection of legal, technical, and business knowledge. For example, identifying white spots in the patent landscape or shaping a protection strategy for AI-driven products requires collaboration across domains.

In short, the IP professional of the future will be part strategist, part analyst, and part communicator—with AI as a strong partner in the background. At thyssenkrupp, we see this as an opportunity to elevate the IP function and actively shape innovation, rather than just protect it.

"[...] from 'AI as an experiment' to 'AI as a capability.'"

In your view, what distinguishes a successful AI implementation in IP management from a merely experimental one? What critical success factors have you identified?

A successful AI implementation in IP management goes far beyond just running a pilot or testing a new tool. The key difference lies in embedding AI into existing workflows in a way that delivers measurable value and aligns with strategic goals.

At thyssenkrupp, we’ve found that three critical success factors are essential:

1. The use case must be clearly defined and relevant. Whether it's improving invention disclosures, accelerating prior art searches, or identifying white spaces, the AI solution must address a concrete challenge within the IP lifecycle. Abstract experimentation rarely scales.

2. Data quality and accessibility are crucial. AI systems are only as good as the data they are trained on. In our case, this includes internal invention records, structured patent data, and even broader contextual data—from supply chain behavior to technology trends. Ensuring clean, well-structured input data is often half the battle.

3. Interdisciplinary collaboration is vital. Many of our successes have come from close interaction between IP professionals, data scientists, engineers, and business leaders. This helps ensure that AI is not just technically feasible, but also legally sound and commercially meaningful.

Ultimately, it’s about shifting the mindset—from "AI as an experiment" to "AI as a capability." When AI becomes part of how we think and work every day, that’s when it really starts to deliver long-term impact.

Could you walk us through specific AI-driven use cases currently in operation at thyssenkrupp—for example, in patent monitoring, portfolio optimization, or prior art analysis?

Certainly. At thyssenkrupp, we’ve implemented a range of AI-driven use cases that support different stages of the IP lifecycle—always with the goal of enhancing strategic decision-making and operational efficiency.

In the area of prior art analysis, we use AI to scan large volumes of patent literature and technical publications. This enables us to identify relevant prior art much faster and with greater accuracy than conventional methods. These tools help reduce manual workload and improve the quality of novelty assessments, especially during invention evaluations.

For patent monitoring and competitive intelligence, AI systems help us track new filings across technologies and regions in real time. By analyzing patterns in competitor filings or detecting emerging technology clusters, we can react quickly—whether by adjusting our own patent strategy or identifying potential infringement risks early on.

Portfolio optimization is another important field. With the help of AI, we can cluster patents, detect overlaps or gaps, and simulate potential future developments. This allows us to make data-driven decisions on where to invest in new filings, where to divest, and how to align our portfolio more closely with business strategy.

Beyond these core functions, we also apply AI in invention workshops, where it supports creativity techniques like SCAMPER, technical problem solving methods like TRIZ or by surfacing unexpected analogies or prior solutions. AI can also facilitate structured interviews with inventors, helping to complete invention disclosure forms faster and more thoroughly.

Each of these examples demonstrates how AI serves as both an accelerator and an amplifier of human expertise—never replacing our IP professionals, but helping them to be faster, more precise, and more strategic in their work.

How do you ensure that AI tools used in IP tasks—such as prior art searching or patent categorization—produce accurate and reliable results?

Ensuring the accuracy and reliability of AI tools in IP tasks is absolutely essential—especially when the results directly influence legal or strategic decisions.

At thyssenkrupp, we take a "human-in-the-loop" approach. This means that AI tools support and accelerate our processes, but the final evaluation always remains with experienced IP professionals. For example, when using AI for prior art searches or patent categorization, the system might propose relevant documents or classifications, but an expert will verify and interpret the findings before any conclusions are drawn.

We also validate AI models regularly. For instance, we benchmark AI-generated search results against expert-curated results to measure precision and recall. This helps us fine-tune the tools over time and build confidence in their outputs.

Another important factor is context awareness. Our IP landscape is complex and diverse—ranging from industrial AI applications to chemical engineering. That’s why we aim to use domain-specific models or tailor general-purpose tools to our needs, ensuring better relevance and fewer false positives.

In short, while AI significantly enhances speed and scale, we never take results at face value. The combination of advanced tools and professional judgment is what ensures reliability—and ultimately, high-quality IP decisions.

What challenges did you encounter during the deployment of AI technologies in your IP processes, particularly regarding data quality, training, or integration with existing workflows?

Like any meaningful transformation, deploying AI in IP processes comes with its own set of challenges—especially in a complex, interdisciplinary environment like ours.

Data quality and availability were among the first hurdles. IP-related data comes in many forms—structured and unstructured, internal and external, historical and real-time. For example, prior art searches often require integrating patent databases with scientific literature or internal invention records. Ensuring consistency, completeness, and interoperability across these data sources took time and cross-functional effort.

Another important challenge was training our IP professionals to effectively use AI tools in their day-to-day work. While the tools are powerful, they’re only as useful as the people who apply them. This meant investing in training programs and practical guidance to help colleagues understand what AI can do, how to interpret its outputs, and where human judgment remains essential. Bridging the gap between legal expertise and data-driven insights required a shift in mindset—and that doesn’t happen overnight. It takes time and repeated exposure to build confidence in using AI as a strategic assistant.

Integration into existing workflows is also a key focus. AI tools are only valuable if they’re actually used. So we spent considerable time embedding AI outputs into the tools and platforms that our teams already work with—whether it’s patent databases, disclosure systems, or dashboards. Change management is critical here. We have to build trust in the tools, demonstrate value through pilot projects, and ensure that the AI augmented existing processes rather than disrupting them.

Lastly, there’s always a balance to strike between automation and control. We are careful not to over-automate decisions that have strategic or legal consequences. The goal is to enhance human judgment, not replace it.

These challenges are valuable learning experiences. They help us to develop not just technical solutions, but a shared mindset for working with AI across disciplines.

To what extent have AI systems helped in identifying white spaces or emerging technology trends within thyssenkrupp’s innovation pipeline?

AI has become a valuable ally in identifying white spaces and emerging technology trends within thyssenkrupp’s innovation pipeline. By analyzing large volumes of patent data, scientific literature, and internal innovation records, AI systems can highlight areas that are underrepresented in the current IP landscape or gaining momentum in specific industries.

For example, by clustering existing patent portfolios and mapping them against competitors and global trends, AI helps us spot gaps in protection—white spaces—where there’s room for strategic expansion. These insights support our business units in proactively steering their R&D efforts toward areas with high innovation potential and low IP saturation.

In parallel, AI-driven trend analysis enables us to monitor technological developments across industries, including signals from startups, academia, and suppliers. This helps us anticipate shifts in the market and align our IP strategy accordingly—whether it’s prioritizing filings in a certain domain or rebalancing portfolio investments.

These capabilities are especially useful in interdisciplinary fields, like digitalization and industrial AI, where traditional boundaries between technologies are blurring. AI helps us detect unexpected connections and emerging cross-domain applications that might otherwise go unnoticed.

That said, we always combine these insights with expert interpretation. While AI reveals patterns and anomalies, it's the combination of data-driven foresight and business context that ultimately drives our strategic decisions.

How do you assess the risks and limitations of relying on AI in high-stakes IP decisions? Are there areas where human judgment remains irreplaceable?

We’re very conscious of the fact that while AI brings significant benefits, it also has limitations—especially when it comes to high-stakes IP decisions. That’s why our approach at thyssenkrupp is always centered on augmenting human expertise, not replacing it.

AI systems are excellent at processing large volumes of data, detecting patterns, and providing initial insights. But IP decisions often involve strategic nuances, legal interpretation, and business context that go far beyond what an algorithm can understand. Questions like: Is this invention aligned with our long-term portfolio strategy? What competitive signal does a certain filing send? How might a court interpret this claim wording?—these require human judgment, experience, and sometimes even intuition.

One of the key risks we manage is over-reliance on AI outputs without sufficient validation. We mitigate this by always keeping a human in the loop, especially when it comes to final decisions on filings, oppositions, or portfolio realignments. We also make sure that our teams are trained to critically assess AI results—looking not just at what the system outputs, but also how and why.

Another limitation is bias in the underlying data. If the training data reflects historical blind spots—such as underrepresented technologies or geographies—AI may replicate those biases in its recommendations. That's why diverse and regularly updated data sources are essential.

In short, AI helps us work smarter and faster, but human expertise remains irreplaceable—especially where decisions have long-term legal, financial, or strategic impact. We view AI as a partner that supports our experts in making better-informed choices—not as a decision-maker itself.

With the pace of technological development, how do you anticipate the role of AI in IP management evolving over the next 5 to 10 years—especially for innovation-heavy industries like yours?

Over the next 5 to 10 years, we expect the role of AI in IP management to grow significantly—particularly in innovation-driven industries like ours, where the speed and complexity of technological development continue to increase.

One major shift will be toward more predictive and prescriptive applications. Today, AI helps us analyze what has happened and what is currently happening in the IP landscape. In the future, we anticipate that AI will increasingly support forward-looking decisions—for example, by forecasting which technologies are likely to become strategically important, or by simulating different IP strategy scenarios based on evolving business goals and market conditions.

We also expect further advances in semantic understanding and contextualization. AI tools will become better at understanding not just keywords, but the intent and technical substance behind patent texts, invention disclosures, and research papers. This could dramatically enhance tasks like claim drafting, portfolio mapping, or white space analysis—making them faster and more robust.

Another area of evolution is integration. AI will no longer be seen as a separate layer, but as a native component of enterprise platforms—tightly embedded into invention workflows, portfolio dashboards, and even contract or licensing management tools.

That said, as AI systems grow more powerful, governance, explainability, and ethics will become increasingly important. Especially in high-stakes decisions, companies will need to maintain transparency about how AI influences outcomes and ensure alignment with legal and regulatory standards—such as the EU AI Act or Cyber Resilience Act.

For us at thyssenkrupp, the vision is clear: AI will continue to empower IP teams to act faster, see further, and think more strategically—but always in collaboration with human judgment and business expertise. The companies that master this symbiosis will shape the next era of innovation leadership.

Are there particular AI capabilities you believe the IP tech industry still needs to develop to unlock the full potential of intelligent IP management?

Yes, while AI has already added tremendous value in IP management, there are still several capabilities that need to evolve for us to truly unlock its full potential.

One area is context-aware reasoning. Current AI systems are good at surface-level pattern recognition, but they often struggle to understand technical depth, legal nuance, or commercial relevance in a holistic way. For example, distinguishing between a truly inventive concept and a marginal improvement still requires human insight. We need AI that can better interpret the why behind a filing—not just the what.

Another critical need is for multimodal data processing. In innovation-heavy industries like ours, valuable insights are scattered across technical drawings, lab reports, sensor data, emails, and patent texts. Future AI tools should be able to integrate and analyze all these different formats together, rather than treating them in silos. That’s especially important for use cases like white space analysis or early-stage invention support.

We also see a gap in intelligent automation of drafting and prosecution support. While there are some promising tools for claim generation or office action responses, they often lack the reliability or domain customization needed for industrial-grade use. There’s potential here for systems that truly support attorneys and agents throughout the lifecycle—from drafting to grant to opposition.

And lastly, we believe the industry needs better solutions for AI explainability and auditability. As AI takes on a more strategic role, decision-makers need to trust—not just use—AI outputs. This means being able to trace conclusions, understand the rationale behind recommendations, and ensure compliance with evolving regulatory frameworks.

At thyssenkrupp, we’re optimistic about the road ahead. But unlocking the full potential of AI in IP will require close collaboration between legal experts, technologists, and software providers—always grounded in real-world needs.

"[...] AI is not a luxury—it’s becoming a necessity."

What advice would you offer to other corporate IP leaders who are considering adopting AI-powered solutions within their departments?

My advice to other corporate IP leaders would be: start with purpose, not with technology. The most effective AI initiatives in IP aren’t those that simply follow a trend—they’re the ones that solve real, clearly defined problems within the organization. So begin by identifying the pain points in your current processes—whether it's the speed of invention disclosures, prior art research, competitive monitoring, or portfolio alignment—and then explore how AI can help address them.

Second, treat AI as a strategic enabler, not just an efficiency tool. Yes, it can automate routine tasks, but the real value comes from empowering your teams to make smarter, faster, and more forward-looking decisions. That requires embedding AI into your IP strategy, not bolting it on.

Third, invest in your people. Technology adoption only succeeds if your team trusts and understands the tools. Focus on building digital literacy and creating a culture where AI is seen as a partner, not a threat. At thyssenkrupp, for example, we’ve found that cross-disciplinary training and collaboration between IP professionals, engineers, and data experts is key to making the most of these tools.

Finally, start small, but think big. Pilot focused use cases, measure impact, learn fast, and scale what works. And don’t be afraid to evolve your internal processes along the way—because integrating AI often means rethinking how IP work is done.

In a world where innovation cycles are accelerating, AI is not a luxury—it’s becoming a necessity. But it’s not about replacing judgment; it’s about enhancing strategic IP leadership in a data-driven age.

Stephan, thank you very much for the interview!

About the interviewee:

Stephan Ising is since 2020 the Head of IP Strategy and Research with the thyssenkrupp Intellectual Property GmbH which bundles the IP staff of the thyssenkrupp group. Prior to that, he worked as Expert IP Strategy and Research for thyssenkrupp IP. thyssenkrupp IP uses AI regularly for its patent work. Stephan is also member of an industry working group on AI and IP. He has a Master in IP Law and Management and a Master of Science in industrial engineering.

At the AI & IP FORUM at Munich on 8 July, Stephan is going to moderate the Use of AI in IP Management panel and he is going to be on the final panel on the present and the future of AI in IP.