The AI shift

Transforming Transport and Communication Museums.

48th IATM Conference from 14 to 17 September in Germany

From 14-17 September 2025, IATM members and those interested in museums and AI gathered in Frankfurt am Main and Mannheim for the 48th IATM Conference! “The AI Shift: Transforming Transport and Communication Museums” examined the potential and challenges of artificial intelligence in transport and communication museums.

Feedback

We are interested in your opinion on IATM’s 48th Conference and your suggestions for future events. Whether you attended or not: Please submit your impressions, and ideas to us via the Feedback Form below!

What happened at "The AI Shift"?

Held under the heading “The AI Shift – Transforming Transport and Communication Museums“, IATM’s 48th Conference took place in Germany from 14-17 September 2025 in the cities of Frankfurt am Main, Mannheim, and Zeppelinheim. It was co-organised by a team from the Museum for Communication Frankfurt, the TECHNOSEUM in Mannheim, and the Deutsche Technikmuseum in Berlin. The theme “The AI Shift” was explored through two inspiring keynote speeches, eight engaging workshops, and lively discussion rounds. In addition, two excursions allowed participants to dive deeper into the world of museum collections and exhibitions — visits to the Museum for Communication Frankfurt’s large depot in Heusenstamm and the Zeppelin-Museum in Zepplinheim complemented the conference with fascinating insights.

14 September 2025

Our pre-conference excursion took us to one of the depot’s of the Museum Foundation for Post and Telecommunication in Heusenstamm. With around 15,000 square metres, the storage facility is the largest depot site of the museum foundation.
For this reason, all large-scale objects of the foundation are stored here. In addition to mail coaches and motor vehicles, this includes mail sorting systems and other large objects from letter and parcel logistics. Located in close proximity to Frankfurt, the site now brings together all parts of the collection that were previously spread across six different locations in and around the city. Today, the depot primarily houses collections related to electrical and electronic communication: telegraphy, telephony, radio, broadcasting, television, as well as computers and the internet.

As part of a guided tour, the depot staff presented the exceptionally impressive object collection to the conference participants and exchanged ideas and insights in response to their questions. Following the day’s program, all participants came together for a shared dinner with local specialties at an Apfelwein-restaurant in Frankfurt. Over schnitzel and green sauce, participants exchanged their experiences and got into the spirit for the upcoming conference.

15 September 2025

On the following day the conference opened with a warm welcome by the president of IATM and director of the Deutsche Technikmuseum, Joachim Breuninger, and the director of the Museum for Communication Frankfurt, Dr Annabelle Hornung.

In his opening keynote, science reporter and expert on AI in radio broadcasting, Ralf Krauter, explored Artificial Intelligence: How AI Will Change Our Life and the Future of Museums through a variety of real-world examples. Starting with a summary of The AI Revolution with key developments from DeepBlue winning a chess tournament to the publication of Chat-GPT and subsequent advances in the technology, he paved the way for an overview of its implications for business and society with predictions on the changing work environment and tasks to be expected. Additionally, he explained the meaning and characteristics of AI in a brief overview. Through recent examples of museums applying AI, our participants could gain a broad impression of what is happening in different departments at varying institutions. This concerned collection management, user experiences, and streamlining operations. His talk continued on risks and challenges before concluding in a Museum AI Vision for 2030.

He outlined four key perspectives on AI and the future of museums:

Focus on purpose: Align AI initiatives with the museum’s core mission and values, and identify promising use cases.
Strengthen data foundations: Invest in digital infrastructure, data quality, and staff skills.
Adopt an ethical framework: Ensure responsible, ethical, and inclusive AI practices. Value transparency.
Interact & collaborate: Use AI to improve operational efficiency and create personalized learning experiences that enhance relevance and visitor engagement.

1st WORKSHOP SESSION

Following the introductory keynote, a workshop session featured four parallel workshops, each with a thematic focus on worries, risks, and legal questions surrounding the use of AI.

In the first of four parallel workshops, Dr Justus Dreyling (Policy Director, COMMUNIA) and Dr Meredith Greiling (Principal Curator of Technology, National Museums Scotland) addressed the theme “Museums, AI, and Copyright.”
For decades, museum professionals have championed free and open online access to collections, including open access to research based on these holdings. Twenty years ago, the sector debated the loss of control inherent in allowing public sharing of collection data and images. Today, the rise of AI raises new questions: Does AI’s use of digital assets risk public trust, institutional reputation, and revenue streams? And can museums effectively address misuse or misinformation?

This workshop explored how museums can mitigate risks associated with AI’s use of their collections, and whether copyright or other intellectual property rights (IPR) tools offer viable solutions. Participants discussed three central conditions for navigating this landscape, based on a report by DACS (Digital Administration and Cyber Security), ‘AI & Artists’:

- Compensation for the use of museum assets,
- Control over how these assets are used, and
- Credit for the institutions and creators behind them.

The discussion underscored the need for proactive strategies to balance openness with protection in the age of AI.

Together with Dr Ko Watanabe (Senior Researcher at Innovation & Quality Center and Smart Data & Knowledge Services, German Research Center for Artificial Intelligence) participants examined the concept of “Trustworthy AI.” A critical consideration as museums increasingly integrate AI into operations and visitor experiences. The session explored the core principles of trustworthy AI, emphasizing transparency, fairness, accountability, and ethical practice as the bedrock for responsible AI adoption.

Through museum-specific case studies—spanning collections, exhibition narratives, and visitor services—the workshop addressed unique challenges. Among them personalization, data quality, and the issue of AI “hallucination” or confabulation – the problem of AI generating responses or content presented as fact that are false or misleading. Participants discussed practical strategies for designing and evaluating AI systems in line with ethical responsibilities, aiming to enhance compatibility between AI and museum contexts.
Key findings from the discussion included:

Many participants encountered the term “Trustworthy AI” for the first time.
Significant age-related gaps exist in AI domain knowledge.
While AI is recognized as powerful, its trustworthiness remains a central concern.

In his workshop, Frank Meister (Head of Digital and IT, visitBerlin, Berlin’s agency for Tourism and Congress (Berlin Tourismus und Kongress GmbH)) posed the critical question: “How do we stay visible in the AI world?” (PDF) As technological progress accelerates, Artificial Intelligence and Applied Machine Learning in particular, delivers significant user benefits and efficiency gains. Yet, these advances also demand transformation to maintain visibility amid rapid change.

The session explored the framework conditions, opportunities, risks, and limitations of AI, with a focus on the tourism sector. Key discussion points highlighted that:

User interfaces are evolving, requiring new approaches to engagement.
Steering visibility will differ from today’s Google-dominated landscape.
Publishing data presents a vital opportunity for organizations to remain visible in the AI era.

The fourth workshop addressed a pressing question: “How can museums responsibly integrate generative AI while preserving authenticity and public trust?“

Tuuli Ahlholm (Digital Heritage Expert, Finnish Postal Museum) presented the recently drafted AI guidelines of The Finnish Museum of Games in Tampere, designed to help staff balance AI’s utility with its ethical challenges—from environmental costs to the risk of devaluing original ideas.

The session fostered open discussion on maintaining curiosity and criticality amid rapid AI development, inviting participants to share perspectives, challenges, and strategies for future-proofing museum practice.
Key takeaways from participants included:

Authenticity and trustworthiness are top priorities.
Guidelines for AI are urgently needed, yet almost no institutions have them.
Avoid using AI in ways you cannot control.

This intense session concluded in a lunch break and tours through the Museum for Communication Frankfurt. Located along Frankfurt am Main’s Museum Embankment, the museum features the permanent exhibition “Media (his)stories retold“. It provides a comprehensive insight into the history of human communication as well as current developments in the field. Additionally, the temporary exhibitions focus on different matters of communication. During the IATM conference, they were dedicated to the exhibitions “New Realities: Fashion Fakes – AI Factories” and “Speaking of Sex”.

2nd WORKSHOP SESSION

In the second round of workshops, participants explored the opportunities of using AI as part of four additional parallel sessions.

At the forefront of technology and cultural heritage, the Swiss Museum of Transport launched a groundbreaking temporary pilot project in autumn 2024. The centerpiece was an AI guide allowing visitors to engage in real-time conversations with digital avatars, exploring the history of the DS Rigi—the world’s oldest preserved side-wheel paddle steamer and a museum highlight since 1959.

Visitor feedback revealed the AI’s ability to host dynamic, interactive dialogues of varying length and depth, adapting to individual preferences and language needs. The project demonstrated that AI content works, but the customer journey remains a challenge.
In the workshop, Martin Ettlinger (Chief Business Development and Market Officer, Swiss Museum of Transport) led discussions on key insights and future applications, focusing on:

Moving beyond QR codes to screens and holograms for seamless interaction.
The evolving role of objects and moderators in visitor engagement.
Public comfort with conversing with machines and the importance of easy language in AI.

As museums increasingly adopt AI in exhibitions, collections, and visitor engagement, the workshop “Making Sense of the Machine: Why Explainable AI Matters for Museums” lead by Matthias Stier (Department Head Digital Strategy, German Museum of Technology) explored a key question: Can we understand how AI systems work — and explain them to others? The session highlighted the importance of Explainable AI (XAI) in the cultural sector, where transparency, trust, and public accountability are essential. Unlike in purely commercial or technical fields, museums must critically consider the societal and ethical implications of using opaque “black-box” algorithms.
Through a short input and an open discussion, participants examined the cultural dimensions of AI transparency: How can explainable AI foster ethical decision-making, inclusive storytelling, and digital responsibility? A central takeaway emerged:

Explainable AI is not (yet) explainable.

Participants emphasized the need for a critical approach to AI, encouraging open dialogue, critical questioning, and ongoing reflection on how museums can responsibly integrate such technologies into their mission.

Dave Thomas (working as Chief Information and Digital Transformation advisor to several institutions) invited action with “Getting started with the best AI Use Case.“ He invited all conference attendees to respond to several questions via a Mentimeter (PDF). Drawing from the responses, he and the participants explored how museums can effectively get started with AI. They identified the most impactful and achievable use cases — and recognized areas where caution or inaction may be more appropriate. Drawing on recent developments at the London Transport Museum, the session highlighted concrete strategies for prioritizing AI initiatives within the cultural sector.

With over 20 years of experience applying AI in climate science, museums, and collections — including work on image identification, collections metadata, and predicting visitor movement patterns — Dave Thomas shared insights on combining UK library data, open data policies, and institutional collaborations. The participants shared how they currently use AI, their main concerns, and their hopes for the future.

Key takeaways included:

Investment is essential — in both time and money.
Prioritize based on impact and achievability.
Don’t be afraid — consider the broader history of technology.

The session encouraged a proactive yet thoughtful approach to AI adoption in museums, grounded in critical reflection and practical action.

In a joint workshop Dr Annabelle Hornung (Director, Museum for Communication Frankfurt) and Magnus Restofte (Director at ENIGMA – Museum of Communication Denmark) discussed our “New Realities“ of AI in and for exhibitions and museum programmes. Playful insights into the range of applications paved the way for a lively discussion. Participants particularly focused on using AI, public interest, and education. Their ideas resulted in several suggestions:

Integrate AI into your digital strategy – we are working with AI and AI is working with us, label AI
Be aware of and act in service of the public, e.g. collaborate with universities, develop a “Code of Conduct”, be aware of “historical evolution”
Educate about AI with a focus on multiple perspectives, explanations, and sustainability (Carbon Policies)
Will AI transform the way we work? How do we respond?

After the workshop findings were shared and discussed during a panel, everyone freshened up before starting into IATM’s General Assembly. The Board reported on the activities of the last year and members shared their recent news. Two new members were elected to the IATM board; Elizabeth McKay, Director and CEO of London Transport Museum and Jens Toftgaard. Director of Danmarks Jernbanemuseum Odense.

A brief outlook towards next year’s conference in Belfast will be followed up by a Save The Date soon. Please find the Meeting Minutes here (PDF).

The busy and inspiring first day concluded with dinner at the Museum.

16 September 2025

The early morning greeted us with a bus ride to the TECHNOSEUM Mannheim. The TECHNOSEUM – State Museum of Technology and Work in Mannheim was founded in 1985 and is one of the three major technology museums in Germany. It presents the history of technology, society, and industry in southwestern Germany from the 18th century to the present day. With its live demonstrations and hands-on activities, it holds a unique position across Germany and is particularly popular with families. With over 160,000 visitors annually and around 35,000 young people taking part in its diverse STEM education programs (Science, Technology, Engineering, and Mathematics), the museum is a leading extracurricular learning venue in the Rhine-Neckar metropolitan region.

The institution was introduced by Claudia Paul (Head Public Affairs & Events, TECHNOSEUM Mannheim) and Dr Alexander Sigelen (Head of Collections, TECHNOSEUM Mannheim) and through subsequent guided tours of the grounds.

With his keynote “A Third Party? The Transformative Impact of AI on Museums” Prof Johannes C. Bernhardt (Professor for History and Interactive Exhibition Design at University of Konstanz) set the tone for the day.

Starting with a quote by George Westerman: “When it comes to digital transformation, digital is not the answer. Transformation is,” Johannes Bernhardt highlighted a new triangular constellation that is emerging between teams, communities, and the new “third player” AI. All three will influence and shape the work of museums in the future. At the same time, he warned that just because a technology appears intelligent does not mean it truly is. AI systems act rationally. It always depends on data. Technology should not be used simply because it exists. Rather, museums should focus on their own challenges and topics. Subsequently, rethinking the processes — a principle that applies to the entire organisation, AI is a leadership responsibility.

WOLRD CAFÉ SESSION

Afterwards, Board Member and CEO of London Transport Museum, Elizabeth McKay, invited participants to an active World Café Session. Grouped around five tables different examples and matters around AI sparked lively discussions and cross-inspiration.

At the table led by Ute Dotti (Chief Exhibition and Collection, Swiss Museum of Transport), she shared the museum’s approach in applying AI for visitor engagement. It was the foundation to a critical reflection on AI’s role in our work and particularly museum education: Participants were split between wanting to use AI to enhance visitor experiences and engagement by providing low-level entry ways while at the same time working on demystifying AI to educate about its potential and dangers.

This was very much on the same note as the table headed with “Why we fAIled so far.” Here,
Dr Tabea Widmann (Digital Education, Zeppelin Museum Friedrichshafen) drew from her recent experiences in developing the exhibition “Image and Power” which focuses on media literacy and digital competences. How do we use, explain, and navigate with AI?

While some reluctancy is understandable and part of dealing with new technologies, they come along with many advantages. Jana Deutschländer (Managing Director, Solvatec GmbH) invited a forward-thinking approach when museum documentation meets AI. She highlighted the opportunities of the powerful tools for automating inventory management and creating new access to collections. The participants shared strategies for the sustainable use of AI and reflected on interoperability, ethical concerns, and encouraged each other to shape meaningful human–AI collaboration.

Just a little bit of history repeating? Dr Martin Weiß (Curator at the TECHNOSEUM) led a discussion on the “History and breakthrough of AI.” Together, the participants explored how a historical perspective can help frame the discussion on the impact of AI. By drawing from pivotal events in the history of AI, the very broad and heterogeneous field of AI was explored. This built the foundation for a critical and open approach towards the technology and the tasks of museums at hand: What historical items should we aim to preserve? Can museums, with their collections and through their public reach, offer a unique contribution to the way in which we approach and develop this new technology?

Closely connected to these topics, was Michael Falser (Head of Research Institute at Vienna Museum of Science and Technology) with his table round on “Artificial Intelligence! Archival Turn in Museums?” By sharing current case studies of AI-supported applications in archival research in the Vienna Museum of Science and Technology, he opened the floor for critical discussions.

The final conference dinner was held on the banks of the Rhein river at the Rheinterrassen Restaurant.

17 September 2025

Our final excursion and study day introduced us to “Zeppelinheim,” a residential area built in 1936 for the flight and ground crews working at the nearby “Flug- und Luftschiffhafen Rhein-Main“ (Airport Rhein-Main) where transatlantic voyages by Zeppelins took off. Today it is a suburb of the city of Neu-Isenburg and the Zeppelin-Museum is dedicated to the history of Zeppelins.

The building’s architecture reflects on Zeppelins and leaves enough room to gain an impression of the dimensions. In addition to an introduction to Zeppelinheim’s history, we learned a lot about the development of the Zeppelin and airship technology and were impressed by the large scale models of different Zeppelins. A visit to the Zeppelin-Museum is recommended. 😉

We are looking forward to seeing each other again or for the first time in Belfast, Ireland in September 2026!

Report by Stefanie Rother, Anne-Sophie Gutsche, and AI, edited by Meredith Greiling.

Downloads

Conference Programme

Here’s an overview of the Conference Programme from 14-16 September 2025 in Frankfurt am Main and Mannheim, Germany

Conference Reader

Please have a look at our Conference Reader for a more in-depth overview of the talks and workshops at the 48th IATM Conference.

Presentations

Several speakers shared their presentations or results with us. Please see here:

Frank Meister on Staying Visible in Times of AI
Tuuli Ahlholm on the AI Guidelines developed at the Finnish Museum of Games
Dave Thomas collected answers on attitudes towards AI

Conference Report

You can download the full report from IATM’s 48th Conference as a PDF here.