On September 25–26, 2025, researchers from the University of Tübingen and the University of North Carolina at Chapel Hill (UNC–Chapel Hill) came together for the transatlantic, interdisciplinary workshop "Aligning AI with Society". Around 50 scholars from computer science, ethics, media studies, philosophy among others worked on site in Chapel Hill as well as online on key questions of AI alignment and the societal impacts of artificial intelligence (AI).

The workshop was organized by Jana Hecktor, Lisa Koeritz, Prof. Jessica Heesen, and Dr. Mone Spindler from the International Center for Ethics in the Sciences and Humanities (IZEW) at the University of Tübingen, as well as by Prof. Snigdha Chaturvedi, Prof. Danielle Szafir, Anneliese Brei, Amartya Banerjee, and the Global Office of the University of North Carolina.

The workshop was made possible through funding from the joint Seed Fund for Research and Emerging Technologies of UNC and the University of Tübingen.

Focus of the Workshop

The event focused on the question of aligning AI systems with societal expectations and perspectives, which has become increasingly relevant in recent years. To approach the topic, one of the aims of the workshop was to give participants insight into the perspectives of other researchers and their disciplinary perspectives on the topic, and to facilitate mutual exchange about their respective research. One of the key findings of the workshop was that researchers from computer science and the humanities use the same terms in many areas but often apply different definitions to these terms. On the one hand, this can lead to challenges in scientific communication, but on the other hand, it may also shape interdisciplinary collaboration. It is therefore crucial to be aware of these differences in order to enable the most productive collaboration possible between the various disciplines. This perspective was used during the event to discuss the various discipline-based questions about the possible, necessary, or debatable alignment of AI.

Workshop Program

For a detailed overview, the full workshop program is available online here.

Day 1

The first day began with a welcome by the organizers, followed by the opening keynote by Prof. Neil Gaikwad from the UNC School of Data Science and Society. His keynote, titled “Aligning AI with Society,” provided a wide-ranging introduction to the workshop’s topic. Neil Gaikwad discussed the role of AI alignment, including the tensions between those who design AI technologies, the groups for whom these technologies are ultimately built, and the broader societal implications of these design decisions.

Afterwards, participants had the opportunity to get to know each other better and talk about their respective backgrounds in a “networking speed dating” session. Building on this initial exchange, a poster session followed, during which there was time to learn about the research of the other participants. Since the researchers came from very different disciplines, very lively discussions ensued. 

He emphasized that the alignment of technology and society is very important, but that the dimension of sustainable development in AI research must not be ignored and implications such as the enormous energy consumption of AI must be taken seriously as a challenge. 

Another important part of the event was the subsequent “flash discussions,” in which participants gathered in groups to discuss their own perceptions of AI and “AI alignment.” The groups were designed so that researchers from different disciplines could work together in small  group to enable the broadest possible exchange of ideas. Among the questions discussed were: Can AI alignment be designed in a way that benefits society? To what extent does the vocabulary, theoretical background, or goals regarding AI differ depending on the professional perspective from which the topic is discussed? 

In a joint concluding discussion, the insights from the group discussions were brought together and an initial shared vocabulary for AI alignment was developed.

Day 2

The second day of the workshop began with a keynote by Dr. Katia Schwerzmann from Ruhr University Bochum, who, with her background as a philosopher specializing in media and technology, was able to bring a different perspective to the workshop topic, broadening the existing perspectives. Her keynote was titled “Generative AI’s Normative Rationality: Whose Values and for What Purpose?” and critically examined the question which values are present in current AI debates and the ethical implications of how AI is currently being used in practice. She emphasized that in order to achieve progress in AI alignment close interdisciplinary collaboration is essential, especially to address emerging challenges constructively.

This was followed by a panel discussion in which Dr. Katia Schwerzmann, Prof. Harlin Lee, Prof. Fan Yao, and Dr. Chris Hazard engaged in an intensive exchange on the topic of “The Many Faces of Alignment: Local Practices and Global Challenges.” The discussion was moderated by Prof. Thomas Hofweber, Distinguished Professor at the Department of Philosophy at UNC. The panel offered diverse perspectives on specific questions of AI alignment and led to an intensive and stimulating exchange. The heterogeneous group was able to discuss interesting perspectives on specific questions relating to AI alignment.

The final item on the agenda was a joint discussion in which the shared vocabulary begun the previous day was further expanded and the findings of the workshop were reflected upon. The group also discussed next steps and possible future collaborations before coming together for a final joint lunch.

Reflections and Outlook

The workshop not only facilitated many productive discussions on the days themselves, but also fostered various new connections between the two universities and within the individual institutions. This was particularly due to its interdisciplinary nature. When it comes to using AI responsibly, both a high technical standard and a deep understanding of the ethical and social implications for society are necessary, which shows why it is so important to collaborate in interdisciplinary formats such as our workshop. 

Bringing together the work of researchers from very different disciplines and transatlantic perspectives, and providing researchers with the space to do so, led to lively discussions and provided inspiration for future collaborations. The work of the workshop was continued in early November in Tübingen with a visit from Anneliese Brei from UNC. Her stay provided an opportunity not only to hold further discussions with researchers from the IZEW and to advance plans for future collaboration between the organizers in joint meetings. In addition, on Monday, November 3, the workshop work was continued in a smaller group. Researchers from the University of Tübingen who had participated in the workshop met with Anneliese Brei to reflect on the workshop and continue next steps. During her visit, she also presented her research to the IZEW team and further strengthened the transatlantic research network.

Cooperation Between the Two Universities

We are pleased that, once again this year, events in cooperation between UNC and Eberhard Karls University of Tübingen were made possible through the joint Seed Fund. Established in 2022, this annual funding program has become an important instrument for strengthening cooperation between the two universities—particularly for formats such as this interdisciplinary workshop.

At the end of March 2025, an interdisciplinary delegation from different departments of the University of Tübingen visited the University of North Carolina at Chapel Hill (UNC-Chapel Hill) to work together on the timely and important issues of the intersection between Artificial Intelligence and regulation and come together at a jointly organized conference under the headline “Towards a Transatlantic Roadmap for AI Regulation”.

The conference was organized by researchers Prof. Tori Ekstrand, Prof. Amanda Reid from UNC-Chapel Hill together with Jana Hecktor and Lisa Koeritz from the IZEW at University of Tübingen with funds of the joint seed for research and emerging technologies by both universities.

The delegation from University Tübingen consisted of Prof. Jessica Heesen and Andreas Baur of the IZEW as well as Prof. Tanja Thomas from the Institute of Media Studies, Prof. Bernd Heinrich as well as Tommaso Fia from the Faculty of Law and guest researcher Hüseyin Ates from Social Sciences University of Ankara.

From the outset, there were not only conversations about how to define and concretize these two broad terms, but also what positive and negative effects both phenomena (may) have: Can AI be used to improve society? And what changes, if it is strictly regulated? Does it help us to move into the right directions or are we hindering process and all the important benefits AI can bring to the table? What are the differences in thinking about, working with or regulating AI between the US and the European Union?

These are just some of the questions the group of heterogenous researchers from fields such as media studies, gender studies, law, economics and computer science discussed about together.

Pre-Event

We started the conference having a really interesting discussion with students in the masters seminar “Digital Media & Society: AI Law and Data Privacy” by Prof. Tori Ekstrand in the Transatlantic Masters program of the UNC combining researchers' expertise with curious, engaged and fresh perspectives of students. We reflected on how we can make sure that AI benefits society, what role data plays in the current discourse around AI and how important privacy protection is and can be, especially for vulnerable groups, but also for each and everyone.

Some of these discussions continued as we moved on toward an evening networking event, held at "Top of the Hill”. The evening allowed us not only to foster some overall connections and topics that would play a crucial role in the next two days, but also to get to know each other better and build connections. It was open to everyone involved in and around the event.

Conference Day 1

On Thursday, we started with some more discussion to help identify collective interests, perspectives and challenges regarding the overall topic of AI & Regulation but also to continue connecting people with shared research interests. Based on the interdisciplinary nature of the group of participants, we came together in presorted groups - ranging from “Justice & Society”, “Vulnerable Groups and Discrimination” to “Infrastructures and Policies” and “Media Regulation” -  for the participants to discuss their personal interests in the topic guided by questions such as “Why are regulations important for your specific research field? Especially today and with regards to AI?” or “What are the areas of AI that need to be regulated the most? And how could such a framework look like?”

With the help of the students from the master's seminar, we not only documented the discussions for future usage, but also created mind maps with the most important aspects that were discussed in the different groups, which we then utilized within a collective group discussion. 

Keynote

The day ended with an engaging and insightful keynote talk from Reggie Townsend titled “Responsible innovation begins with responsible innovators”, where he shared his perspective on the topic of “Responsible AI”. He shared his perspective on why we need to use AI for good and be aware of possible pitfalls as well as for the self-regulation of companies with regards to the topic:. Instead of leaning on regulations – which he argued sometimes leads to overregulation, or might stray too far from the reality of creating and using AI systems – we should not overlook companies that hold themselves accountable for generating responsible AI systems. 

For Reggie Townsend, responsibility describes the delicate balance between freedom and obligation. He used this definition to especially talk about “response-ability”. With that, he pointed out the fact, that it is unavoidable that technical systems – even so-called AI systems – can fail. Instead of trying everything to avoid the unavoidable, we should think about how we can best respond to a potential failure, to be prepared. This should be part of our responsible usage of AI and underline the importance of AI governance within research and industry. 

The keynote was followed by a Q&A with host Prof. Tori Ekstrand as well as audience members, which was continued during the reception afterwards. 

Day 2

The second day started with a panel discussion titled “Bridging the Atlantic: AI Accountability During Democratic Decline”. Participants from University of Tübingen as well as from UNC started discussing some of the most prevalent issues regarding AI & Regulation. Shannon McGregor, associate professor and principal investigator UNC’s Center for Information,Technology and Public Life (CITAP), moderated the panel. The group picked up on some of the previously raised issues and questions, but also focused on the unique perspectives everyone was bringing to the table.

Concluding an event is always a challenge. Our aim was to not only look back on the emerging questions, challenges, possibilities that have been discussed as well as the possible outlook of a roadmap toward a transatlantic perspective on AI and regulation, but also to lay the foundation for the following collaborative work. Here, we focused on how to build upon the questions, problems as well as different next approaches we came up with. The group plans to continue this work in the coming months based on the understanding and discussion built within those two days.  
We’re more than looking forward to everything that will come out of this and are sure there is a lot of potential foundation that has been laid during these days.

Thanks!

We want to thank everyone involved. Beginning with Tori Ekstrand, Amanda Reid and Martha Carey, who were amazing co-organizers in putting the conference together. They also made sure that we felt welcome and had a smooth conference on campus. They went above and beyond in fostering interesting discussions and played an integral part in leading them to possible next steps. Thank you all so much!

Also thanks to everyone from UNC and Tübingen who participated and therefore helped to broaden the horizon of perspectives, enriching the discussions and making so many networking moments between both universities as well as over the heterogenous disciplines possible – briefly: Thanks to everyone who helped make this event so fruitful! 

Finally, we want to especially thank the UNC Global Affairs office! For helping us, when help was needed, for welcoming all of us so warmly and friendly and for offering their building not only for the keynote talk, but also for the reception afterwards! Thank you, Melissa McMurray, Krista Northup, Walter Walker Stephenson and everyone else involved in the planning. Thank you also to Katie Lindner from the UNC Center for European Studies.

A special thanks goes to Barbara Stephenson, Vice Provost for Global Affairs and Chief Global Officer at UNC-Chapel Hill, for being so enthusiastic about and encouraging in advancing the strategic partnership between University of Tübingen and UNC-Chapel Hill, but also for welcoming the delegation and for giving welcoming remarks at our keynote talk as well as for our panel discussion! We appreciate all the effort and commitment the Global Office put into this conference.