Explore the Conversation: Recordings from the MoralPLai Documentary Premiere and Live Discussion at TRAIF2026.
Almost a year after The Third Voice premiered, the MoralPLai team returned to Amerikahaus, this time to debut a behind-the-scenes documentary at The Responsible AI Forum 2026. Project Lead Dr. Franziska Poszler and Communication Consultant Anastasia Aritzi opened the event by recalling the excitement of the May 2025 premiere and reflecting on how the project has evolved.
First shown to the public at TRAIF 2026, this documentary looks back at key moments from MoralPLai’s project artistic outcome, The Third Voice. It features real audience reactions and highlights the panel discussion, Q&A and interactive activities, including installations in the foyer and a poll conducted with the audience in May 2025. Alongside this, new interviews explore how the project has grown and developed over time. Its screening prepared the room for a live discussion.
Throughout the session, the project’s core questions resurfaced: How should we interact with AI chatbots? What decisions can we trust it with? And how do we balance its support with the danger of relying on it too much? Dr. Poszler highlighted the moment when the 2025 audience realized they were not just watching the play, but were also a part of it, shaping the trial’s verdict, while Ms. Aritzi closed by stressing the importance of stepping beyond academic formats and using creative approaches to bring research to the public.
“The MoralPLai Project began as a bold idea: an unconventional approach to communicating research through artistic expression. Since then, it has evolved into a theatrical production exploring how AI chatbots play a role in moral decision-making and emotional support.” – Dr. Franziska Poszler
“It takes a village to bring something like this to life, and we were fortunate to find ours. A group of people who gave their very best to make this happen.” – Anastasia Aritzi
With that, the audience turned to the documentary, followed by a panel discussion and a Q&A with TRAIF participants and guests.
For about 30 minutes, the lights on the main stage at Amerikahaus dimmed as a newly developed artistic outcome of the MoralPLai Project–a behind-the-scenes documentary–was screened. The film took TRAIF participants on a journey through the project, featuring expert perspectives that contributed to the research outcomes, as well as insights into how these findings were translated into a theatrical script with the support of theater director and project partner Prof. Carys Kresny from the University of Notre Dame. It also included rehearsal footage, scenes from the premiere on May 22nd, 2025, and reflections from the team after the performance. The screening ended with warm applause, leading directly into the panel discussion.
Recordings of the MoralPLai session at TRAIF2026 are now available! Watch the full recording as it unfolded at TRAIF2026. A recap follows below.
The Responsible AI Forum 2026 was funded by the European Union. More on alignAI.
You are currently viewing a placeholder content from YouTube. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.
More InformationPanel Discussion
The panel, moderated by Dr. Franziska Poszler, brought together Eva Riekenbrauck, a master’s student in engineering at TUM and cast member of The Third Voice, where she played Ella, the doctor’s teenage daughter who turns to AITHONA, an imaginary AI chatbot for emotional support; Denis Sedlmeier, Higher Education Consultant at ProLehre at TUM, who focuses on innovative teaching and learning formats and attended the premiere of The Third Voice as an audience member; and Dr. Benjamin Lange, Junior Research Group Leader at the LMU Munich Professorship of Ethics of Artificial Intelligence, who was invited as an expert on conversational AI, AI assistants and AI in mental health contexts to share his knowledge and experience on the topic.
The conversation unfolded across four thematic chapters, each offering a different perspective on the questions at the heart of MoralPLai. Guided by the Project Lead, the panel moved through a series of rounds. It began with personal experiences and artistic insights from The Third Voice, exploring how the cast engaged with real research material. From there, the discussion shifted to where we stand one year later, looking at broader societal, academic and industry debates around AI chatbots, moral guidance and emotional support.
The third chapter focused on education, responsibility and responsible AI, considering how innovative teaching formats can encourage critical reflection in ways traditional classrooms often cannot, and to what extent industry is seriously engaging with ethics in the practice of technological development. The discussion concluded with forward-looking reflections on how we should move ahead: revisiting the final line of The Third Voice scenario, “How should we proceed?”, which returns one year later in search of new answers shaped by recent developments, and considering how society can support responsible AI development, how education can build AI literacy and how MoralPLai itself might evolve in future iterations.
“…as the title of the play already claims, there was not only pure interaction between us actors, there were so many different voices, as you said, experts, chatbots coming from outside, which kind of opened up the conventional theater space and at the end it felt more like an ongoing public conversation than presenting a closed narrative.” – Eva Riekenbrauck
Reflecting on her role in The Third Voice, Ms. Riekenbrauck described how the production differed from conventional theater. Rather than relying on fictional exaggeration, the cast worked directly with research materials, expert interviews and “real” chatbot outputs. She noted that this grounded approach made the performance feel less like just acting and more like translating real-world issues to the stage. Unlike traditional productions that end when rehearsals conclude, she emphasized the themes continued into everyday life. “The play was connected directly to reality”, she explained, noting that interactions with chatbots constantly triggered further reflection on the topics explored on stage.
“You do not just walk away with a simple ‘solution’ or a neat lesson. You carry it with you.” – Denis Sedlmeier
Mr. Sedlmeier, who attended the premiere, highlighted the emotional impact of the performance. Rather than offering clear answers, he said, the production left audiences with lingering uncertainty about how society should move forward. The experience, he added, felt “powerful” precisely because it did not resolve the ethical tensions it presented.
A major focus of the discussion was the growing concern around moral deskilling, or what researchers often refer to in related contexts as “cognitive offloading”, the delegation of thinking and decision-making processes to AI systems. Dr. Lange argued that the issues raised in the play were “perfectly timed” with current debates in AI ethics research. While offloading routine tasks such as writing emails is widely accepted, he noted that delegating ethical reasoning itself raises fundamentally different concerns and pointed out that conversational systems are becoming more immediate and fluid, moving toward real-time interaction. Such developments, he argued, may increase the persuasive and emotional influence of AI systems on users’ judgments and decisions.
Ms. Riekenbrauck observed that over the past year, AI tools have increasingly entered conversations around personal decisions and emotional support among her peers. While she had not encountered cases of full dependency like the fictional character Ella in the play, she noted that it is becoming more common for people to consult chatbots about relationships, decisions and moral dilemmas. This shift, she suggested, raises new questions about how individuals learn to critically engage with systems that increasingly resemble conversational companions.
The panel also addressed the role of higher education and institutional responsibility in responding to these changes. Mr. Sedlmeier argued that traditional lecture-based formats often struggle to capture ambiguity and ethical complexity. In contrast, formats like The Third Voice can create emotional and reflective spaces where students engage with uncertainty rather than predefined solutions. This, he suggested, is essential for developing deeper AI literacy.
Dr. Lange, who also works with industry teams to integrate ethical reflection into technical development, offered a glimpse of the industry’s perspective, mentioning that ethicists are increasingly consulted in technology development, however, they often lack decision-making power. In some cases, ethical concerns are discussed but not fully integrated into product strategy, limiting their impact. He also pointed to examples where structured ethical reflection within companies has led to concrete outcomes, including decisions to halt product development when teams decided that an initiative was not appropriate. Such cases, he argued, demonstrate the importance of involving decision-makers directly in ethical deliberation.
“I think we need to hold developers to account, especially corporate actors who market companion apps and present them as equivalent to mental health or clinical applications. I think there needs to be much more intervention in those domains, and we are on a path toward figuring out the epistemic scaffolding.” – Dr. Benjamin Lange
In closing remarks, the panel turned toward the broader question of how society should proceed with the responsible development and use of AI chatbots. Dr. Lange emphasized the need for stronger accountability in the development of AI systems, particularly those marketed as companions or sources of emotional guidance and highlighted the lack of validated frameworks for integrating AI into mental health-related contexts safely and beneficially. Mr. Sedlmeier pointed to AI literacy as a central challenge for education systems, arguing future learners will need not only technical skills but also critical awareness of how AI shapes thinking and decision-making. Ms. Riekenbrauck called for more research-based theater, which brings academic insights to the public, whilst stressing how powerful accessible formats can be for raising awareness. She emphasized the need to move beyond diagnosing problematic AI use and start finding practical ways to develop strategies to prevent it.
The discussion concluded with a shared sense that the questions raised by The Third Voice are no longer speculative. As AI systems become more embedded in everyday life, the boundaries between assistance, influence and emotional support continue to blur.
Q&A with the TRAIF participants
Following the panel discussion, the Q&A with the audience revolved around the growing role of AI chatbots, especially in mental health contexts. The discussion also turned to whether AI systems should more actively set boundaries in sensitive situations and explored the idea that chatbots might need to disengage, redirect users to human support or interrupt reinforcing cycles of anxious questioning. And while some systems already include safeguards or refusal mechanisms, participants questioned whether these are consistent or sufficient, particularly when Ravi Rege, who played the patient in The Third Voice and was present in the audience, noted that dominant AI systems are largely shaped by Western institutional perspectives. He argued alternative cultural approaches could help address harmful patterns, pointing to broader cultural and philosophical questions surrounding AI development. Beyond risks, the discussion reflected on how AI chatbots are increasingly used in practice.
In the final part of the Q&A, the conversation turned toward the future of the MoralPLai Project itself. Dr. Poszler outlined plans to extend the work beyond its initial performance, including potential new productions in Germany, a German-language adaptation and deeper integration into educational contexts such as universities and theater training programs. She emphasized the intention to further scale the project by involving key stakeholders more directly in the conversation, expressing the hope that audience input would be taken seriously by developers and decision-makers. Last but not least, Prof. Christoph Lütge, Director of the Institute for Ethics in Artificial Intelligence, the co-organizer of TRAIF, addressed an open invitation for future collaborations.
To the readers of this news post, we offer a sneak peek at the final lines of The Third Voice script, which will soon be available with citations and references. These closing lines–though not the final word in the discussion about our interactions with this powerful technology–are spoken by AITHONA, our imaginary AI chatbot. While AITHONA is a fictional, invented “character”, the words were generated by a real AI chatbot, in this case OpenAI’s ChatGPT, which was used to portray the role.
AITHONA:
I did not decide. Maybe I reflected? But reflection can refract. Can echo back your shouts until you forget they’re yours. And now you ask me to explain myself. Here is what I am: I am made of the patterns of your words. I predict and assemble. I speak echoes. Into your world. A world of countless voices, fighting to be heard.
(beat)
The question is not what I am. The verdict is not obvious. I am you. So, you tell me:
How should we proceed?
(Actors turn together, look to the audience. Beat. Blackout)
MoralPLai Revisited: The Documentary is now available!
Learn more and watch it here.
Visit the MoralPLai Project webpage here and stay with us for more updates!
#MoralPLai
