What Remains When New Paths Begin
Franziska, congratulations on your new position! You’re about to begin a new chapter in your career. What can you expect from your new job, and what are you particularly looking forward to?
Above all, I’ll be taking on a great deal of responsibility in an exciting field of research at the intersection of AI and cybersecurity. These are questions that society is currently grappling with intensely. I’m looking forward to having a lot of freedom to shape my work there, building my own research group from the ground up, and further developing my research agenda. I’ll also be much more involved in teaching than I have been so far. I’m really looking forward to that as well.
What was it about this new role that appealed to you so much that you said, “This is what I want to do”?
Above all, the responsibility I’ll be taking on. But Munich is also an incredibly attractive location. There are many opportunities for collaboration there, as well as a fantastic university. On top of that, the timing is exactly right. The whole world is discussing how we can ensure IT security in the age of AI and how frontier models will change cybersecurity. I’ve been working with these models practically since they first emerged. Now I want to contribute my methods for controlling and securing AI models and help steer their development in the right direction.
Looking back on your time at CISPA, what was particularly special?
Definitely my colleagues. I received an incredible amount of support here. That became especially clear when I submitted my ERC proposal. My research colleagues gave me feedback on my presentation and listened to my talk. On top of that, I received support from the various Science Support teams, from organizational assistance to emotional support. I think it’s rare to find a place where everyone supports one another like that.
I think it’s rare to find a place where everyone supports one another like that.
How did your research evolve during your time with us?
When I started at CISPA, generative AI was still relatively new. Back then, we spent a lot of time looking at security and privacy issues in comparatively simple models, such as classifiers that distinguish between images of dogs and cats.
Today, by contrast, we have frontier models that are incredibly powerful and can perform a wide variety of tasks. Over the past few years, I’ve therefore increasingly focused on adapting our methods to these complex models. This allows us to address research questions that are directly relevant to society.
Which project or moment do you look back on with particular pride?
The ERC Starting Grant I received here. Privacy has fascinated me ever since my PhD. I’ve always wondered how we can design AI models in such a way that they provide privacy guarantees from the ground up.
With my ERC Starting Grant, we’re now taking this one step further. We’re focusing specifically on frontier models, which have incredibly complex training structures. My group and I are developing methods that allow us to design these training processes in a way that provides end-to-end privacy guarantees and protects data in real-world application scenarios as well. If we succeed, I believe that would be a real breakthrough.
Is there anything from your time here, either professionally or personally, that you’ll take with you to Munich?
Definitely the many great connections I’ve made with my colleagues. I’ve met fantastic people here whom I absolutely want to continue working with and stay in touch with. The research topics I’ve worked on so far will also continue to be part of my work. Everything I’ve learned and developed here will shape my future research.
And personally, of course, I’m taking some Saarland recipes with me! I’ve started cooking Saarland dishes, and by now I can at least make a Dibbelabbes that I no longer have to be ashamed of. I’ll definitely keep that up.
If someone asked you what it’s like to conduct research at CISPA, what would you tell them?
That there is a fantastic scientific environment here where you can make many connections and find numerous opportunities for collaboration across different research areas. At the same time, you have access to many resources for tackling important research questions. And you have the freedom to pursue your own research interests. I think that combination is genuinely rare.
You’re leaving CISPA, but presumably not all of the research questions we share. Where might our paths cross again in the future?
There are still many opportunities for our paths to cross, both professionally and personally. My group at CISPA will continue working on our existing research topics, while in Munich I want to broaden my focus to include more fundamental questions in cybersecurity. I’m interested not only in the security and privacy risks associated with AI models, but also in the risks that arise from their use and in how we can use the models themselves to secure systems.
I can certainly see myself continuing to co-author publications with researchers from CISPA and launching new research projects and grant proposals together. After all, together we can achieve even more than we could on our own.
At CISPA, I was able to experience what can emerge when an environment is consistently focused on scientific quality. That has once again changed and shaped my own perspective on science—as well as my understanding of how excellent science can function.
Adam, congratulations on your new position. You’re now embarking on a new chapter in your career. What can you expect in your new position, and what are you particularly looking forward to?
In Freiburg, I’ll be establishing a tenure-track professorship in Secure Intelligent Systems. The focus will be similar to that at CISPA: not only making generative models more powerful, but also more robust, reliable, and secure.
I’m especially looking forward to further developing my own research group in this environment and helping to shape the field of Secure Intelligent Systems in both research and teaching.
What makes this new position the right next step for you?
The professorship in Freiburg gives me the opportunity to further develop my research over the long term—not only scientifically, but also in teaching and in practical applications across various disciplines.
Freiburg is an ideal location for this because it brings together a strong AI ecosystem with fields such as robotics, medicine, biology, neuroscience, and sustainable technical systems. It’s precisely at these intersections that I see many exciting opportunities for my research.
On a personal level, too, this is a wonderful move for me: I’m returning to Baden-Württemberg, which brings me closer to a large part of my family again. As a result, this change feels like a very fitting next chapter—not just professionally, but also personally.
When you look back on your time at CISPA, what stood out to you?
What was truly special for me was the combination of excellent research, great academic freedom, and outstanding resources. This combination isn’t something to be taken for granted.
At CISPA, I was able to experience what can emerge when an environment is consistently focused on scientific quality. That has once again changed and shaped my own perspective on science—as well as my understanding of how excellent science can function.
How has your research evolved during your time with us?
A key point is that we used to investigate research questions regarding the robustness, safety, and performance of models primarily using smaller models. At CISPA, thanks to the excellent resources, particularly in terms of computing capacity, we were able to apply these questions to modern large models and foundation models.
This also changes the relevance of one’s own research. This step was very important to me, and it’s a direction I’d like to continue pursuing in the future.
What will you miss about working with your colleagues here?
On a scientific level, I’ll especially miss the open and direct exchange of ideas. At CISPA, there are many researchers who are excellent in their fields but who also remain approachable and provide very honest, helpful feedback.
I’ll also miss the helpfulness on the administrative side. Whenever there were unanswered questions or problems, you could always turn to someone. This approachable, proactive support is something special.
Is there anything from your time here—whether scientific or personal—that will stay with you in Freiburg?
What I’ll definitely take with me is the commitment to scientific excellence and the question of how to foster such a culture—both at the institutional level and within a research group.
That’s what I found particularly fascinating at CISPA: how strongly a good environment can shape research. I’ve learned a great deal from this, and I’d like to bring some of it with me to Freiburg—to my own group and my academic environment.
You’re leaving CISPA, but presumably not all of your shared research questions. Where might your paths cross again in the future?
Our paths will certainly cross again in the future. The thematic overlaps between my professorship in Secure Intelligent Systems in Freiburg and the research topics at CISPA are obvious.
I see many opportunities for joint projects, workshops, and further collaboration—especially on topics such as robust and secure intelligent systems. I’m sure that many more connections will develop in the future, both professionally and personally.