Master student, KAIST Graduate School of Culture Technology
Pshenova Aleksandra
spshenova@kaist.ac.kr
The museum has carried a strong stereotype that it is entirely past oriented. This traditional concept of the museum comes from how it preserves relics with historical value and passes that value down to future generations. However, in order to stay up to date with the current technological progress, museums will need to undergo the paradigm shift of becoming “digital” while maintaining traditional values.
Innovative digital technologies, such as now trending Artificial Intelligence, or AI, must be adopted to transform museums and the roles they fulfill for the future, while also keeping the museum a relevant and attractive institution for the current digital-driven society. As it might appear, this task is far from easy since it is not easy to accept the changes brought by the digital era in the existing museum system, both to the institutions themself and the public, as museums usually heavily emphasize the value of authenticity of their collections.
In this article, we will try to dig deeper into this complex problem, outlining the history of AI introduction to museums, observing the societal response to the ongoing museum digitization and its challenges, and conclude with a brief brainstorming of the possible solutions to those challenges that could help museums benefit from the newest technology, not resist it. We will also take a look at how museums in Korea address this complex trend of innovative digital technologies penetrating a rather analog domain of cultural heritage.
History of AI implementation in museums
The earliest instance of using Artificial Intelligence (AI) in museums can be traced back to 1970 with the development of Shakey, the first robot that used algorithms similar to modern-day AI. However, Shakey was not specifically created for a museum setting, even though the museum was chosen as one of the testing grounds for the robot. This showed the emerging plans to start applying the most advanced technologies at the time to museums.
When it comes to the implementation of AI specifically in museums, the timeline is more recent, as the robust AI development itself is a relatively recent trend. According to a research study, AI initiatives in museums began to appear around 2010 and have really taken off around 2016. It is difficult to evaluate the exact number of AI initiatives introduced in the past, and even recently, given the broad spectrum and language used to describe such initiatives, but there is no doubt that the museum space has turned into one of the major application frontiers for AI technologies in less than ten years.
<Figure 1> Timeline of AI initiatives in museums (2010 -2018) (N = 61 examples) (Source: French and Villaespesa, 2019)
Museums are increasingly leveraging AI technologies to enhance accessibility and visitor experiences. The Museum of Tomorrow in Rio de Janeiro has been using the IRIS+ chatbot since its opening in 2015 to improve accessibility with real-time sign language translation and audio descriptions for visually impaired visitors, and keeps developing and adding new features to it every year. In 2022, the Louvre introduced Leonardo, an AI-driven virtual assistant for personalized tours. The British Museum launched an Augmented Reality (AR) app in 2023, utilizing AI to create immersive experiences. Additionally, an experimental project called “The Living Museum” was launched in October 2023, demonstrating how AI can be used to engage visitors by allowing them to interact with museum artifacts, though it is not officially affiliated with the British Museum. These initiatives highlight the growing role of AI in transforming museum experiences.
The Current state of AI in the Museum Space
After having explored the history of AI introduction to museums, we can overview the current state of this domain, detailing each major application field of AI in the museum space.
Virtual Museums
Virtual and Augmented Reality (VR and AR) technologies are changing how museums operate by solving space limitations. They allow museums to display digitized artifacts in virtual environments, which require minimal physical space. This makes it possible to showcase a large number of artifacts that would otherwise be difficult to display due to space constraints. As a result, virtual museums enable visitors to experience the museum’s collection in its entirety. Moreover, AI is greatly aiding the digitization process by automation and acceleration, helping restore damaged parts of the artifacts and preserve the highest quality of the digital copies.
Apart from solving the fundamental spatial limitations of museums and the physical inability to showcase all of their collections, VR technologies allow the museum to transform into an immersive and interactive space, attracting new visitors. Immersive technologies like VR and AR enable museums to present complex information in engaging ways, making cultural heritage more accessible and inclusive, thus diversifying the audience. As an example, The Met Unframed, designed by the Metropolitan Museum of New York (USA) in partnership with the multidisciplinary production company Verizon Communications Inc., is an immersive augmented and virtual reality experience available for free on any mobile device. What started as a project to support the museum during the COVID-19 pandemic, now continues to thrive as a portal to the new virtual museum reality.
<Figure 2> The Met Unframed lets you explore one of the world’s greatest museums on your mobile phone (Source: Personal webpage of Jordan Weil, Creative Strategy Lead on the project, January 2021)
Bringing the entire museum space into the virtual world by digitizing all of its collections and creating a digital museum replica allows the museum to transcend the constraints of space and open it to a worldwide audience, fostering a borderless and open knowledge-sharing space. This way, virtualization of the museum trend contributes to the inter-institutional communication between museums and the wider phenomenon of globalization and democratization of knowledge. The Smithsonian Learning Lab is a multidisciplinary online platform that integrates information from multiple Smithsonian institutions and presents it in an engaging and educational way.
Robots in Museums
Robots were first introduced to museums as far back as the 1980s, when industrial robots found their use in artifact handling and helping the museum workers with collection management. In the 1990s, the first robot museum guides, such as MINERVA and RHINO, were first introduced as pioneers of robots developed exclusively for museums, utilizing some basic AI and Machine Learning (ML) components.
One of the earliest examples of AI application in museums dates back to 2016, when a robotic art critic named Berenson was introduced at Paris’ Musée du quai Branly. This AI-powered robot was designed to record visitor reactions and develop its own preferences of certain art pieces.
Nowadays, stationary or mobile robots are a staple in most museums, serving as informational guides or simple entertainment. Examples include the famous Pepper of the Smithsonian, which popularized using robots in museums. The pilot Pepper Robot project was launched in April 2018, when several Pepper models were introduced to four Smithsonian institutions, taking different roles, such as Promoter, Host, or Entertainer (Choreographe). In December 2018, the Smithsonian Organization and Audience Research (SOAR) analyzed the gathered data and visitor studies to publish a comprehensive report covering the overall results of the pilot program.
The report states that in some of the pilot locations, half the visitors who walked in the vicinity of Pepper interacted with it. In other places, two in five visitors reported interacting with Pepper either alone or as a part of a group. Overall, SOAR concluded the report on a positive note, highlighting the potential of the Pepper robot in revolutionizing the educational value of museums by bringing a ‘fun’ and more ‘informal’ aspect to it. Pepper is still considered one of the most popular robotic assistants in the museum space.
This conclusion highlights the potential of using robots in the museum space as all-knowing museum guides or engaging entertainers, creating opportunities for new ways of information delivery and visitor engagement.
<Figure 3> Pepper robot serving as a museum guide for a group of children in Smithsonian’s National Museum of African Art. (Michelle Edwards / National Museum of African Art, 26 April, 2018)
<Figure 4> Pepper experience ratings indicate that more than half the visitors who interacted with Pepper chose the highest satisfaction rating (Source: Smithsonian Organization and Audience Research, December 2018)
Virtual museum guides
Different from virtual museums, virtual museum guides are partially influenced by the AR technology, since they bring virtual reality into the physical museum space. Existing in multiple forms, from audio-only guides that accompany the physical visit, to the actual docent presenting in a form of a virtual avatar and blending into the physical environment while delivering the museum tour, all virtual museum guides serve a single purpose: to engage with the visitor in a variety of unique ways, different from the usual museum experience, and make the museum experience more entertaining, all while staying true to the museum’s ultimate educative role. Empowering the virtual guides with AI helps to elevate the educational value of such visits by leveraging all possible data on a certain artifact to give the visitor a well-informed and complete answer.
Some virtual museum guides utilize cutting-edge technology like Head-Mounted Displays (HMDs) or smart glasses. Others, on the other hand, stick to less robust, but more resource-effective and habitual devices like headphones and smartphones, which can be brought by the visitor or provided by the museum to ensure the stable operation of their local applications.
Recently, a new trend started to gain popularity in the museum domain: the integration of powerful Large Language Models (LLMs), such as ChatGPT, in the museum space. In an attempt to use the capabilities of ChatGPT-4.o to enhance the museum visit experience, a group of researchers led by G. Trichopoulos created a museum audio guide called MAGICAL, which would answer visitor questions and introduce the museum exhibits in the form of an audio guide, generating answers to the visitor’s questions based on the previously learned information about the museum.
AR and interactive exhibitions
We have already discussed AR and VR technologies in the context of virtual museums, but can these technologies complement the physical museum visit experience? There is no doubt that AR technologies, alongside VR, are blending physical and virtual space, creating a unique experience and making cultural artifacts more accessible and understandable, especially the ones that are hard or impossible to access otherwise.
AI-generated interactive narratives, such as games and interactive stories, allow to enrich the visitor experience and attract new audiences, fostering new ways to connect to the artifacts. Contrary to the games that serve a purely entertaining and engaging purpose, some other interactive experiences make use of interactive elements in order to tell stories that were hard to imagine in the past. For example, ‘Mona Lisa: Beyond the Glass’ was a groundbreaking VR experience launched by the Louvre Museum in Paris on October 24, 2019, as part of their landmark Leonardo da Vinci exhibition commemorating the 500th anniversary of the artist’s death.
This immersive VR project, created in collaboration with HTC VIVE Arts, allowed visitors to interact with the world’s most famous painting in virtual space, revealing hidden details and providing insights into Da Vinci’s techniques. The experience offered users a chance to see the Mona Lisa up close, learn about its creation and restoration, and even meet the woman behind the portrait. This project was the Louvre’s first VR project, setting a new standard for unique ways of interacting with art.
<Figure 5> The Mona Lisa in virtual reality in the comfort of your own home (Source: The Louvre Museum homepage, www.louvre.fr, 23 February 2021)
In 2023, the Metropolitan Museum of Art launched ‘Replica’, an innovative partnership with Roblox that transformed the museum experience. This unique app allows visitors to scan artworks at the Met and convert them into digital collectibles for use in Roblox. Users can then explore a virtual version of the Met, customize their avatars with these digitized artifacts, and even curate their own virtual exhibitions.
<Figure 6> Visit the Met and enter the Metaverse: unique experiences provided by museums (Source: The Met homepage, http://www.metmuseum.org, 3 August 2023)
By embracing interactive and immersive technologies, museums like the Met are pioneering new ways for visitors to engage with art collections, bridging the gap between traditional exhibitions and digital platforms.
Preservation and management of museum collections
Apart from the educational and entertainment value of AI-based solutions for the museum space, AI is offering new opportunities for monitoring artifacts and predicting potential deterioration and changes, revolutionizing the field of conservation and restoration.
These opportunities enable museums to find a new use to such tools as VR and AR, which can be used to simulate virtual restoration and demonstrate the impact of different restoration techniques on an artifact even before the artifact is moved to the actual museum space.
For example, the Smithsonian is currently implementing AI algorithms to perform predictive maintenance on its equipment, anticipating breakdowns and addressing them in time. The integration of AI in this case reduces downtime and operating cost of the equipment, ensuring greater efficiency and better operations management.
AI is also revolutionizing visitor studies, efficiently gathering data such as time spent in front of artworks, visitor movement patterns, and interactions with digital installations. Implementing AI into the visitor study domain helps to accelerate data accumulation and analysis, providing valuable insights into possible future attendance trends and preferences of the audience.
This data helps museums better understand audience preferences, as well as to optimize exhibition layouts to improve flow and reduce congestion in the museum space. Immersive experiences created as a result of gathered data analysis help improve visitor engagement and tailor content to visitor preferences. And, most importantly, by analyzing visitor behavior, museums can better safeguard valuable artifacts by managing crow flow and organizing the surrounding space in an optimal way.
Future prospects of AI in museums
The future of AI in museums is poised to transform both visitor engagement and operational efficiency. For example, museums increasingly adopt AI tools to analyze visitor behavior, enabling personalized experiences tailored to individual preferences. Behind the scenes, AI enhances collection management and fosters global collaboration by standardizing databases and streamlining information exchange.
Virtual and immersive platforms are also expected to redefine museum spaces, offering interactive and collaborative experiences through technologies like VR and AR. However, challenges remain, including ethical concerns over AI-generated content, data privacy, and potential job displacement. Addressing these issues through transparency, rigorous testing, and industry-wide collaboration will be essential for realizing the full potential of AI in the cultural heritage domain.
Potential challenges on the path of AI adoption
Having explored the evolution and current landscape of AI in museums, it’s critical to address the challenges inherent in adopting these technologies. While some obstacles vary by institution – such as budget constraints or staff expertise – others are universal and affect the institutions on a global scale.
A primary concern is combating AI hallucinations, which can distort educational content or historical accuracy. In the field of AI, hallucinations are AI responses containing false or misleading information that AI presents as a fact, making it harder for the unprepared users to distinguish from the actual truth. The museum industry is a field that demands factual accuracy, so hallucinations present the biggest challenge and obstacle to widespread AI adoption in museums.
As a result, museums have to rely on their own limited data in order to ensure data accuracy, especially when dealing with visitors that may have a diverse level of background knowledge and understanding. The risk of hallucinations significantly slows down the adoption of AI in museums and the processes involved with it, such as the global integration of museum databases and cross-institution communication. Because museums cannot fully rely on information retrieved from the web, cultural institutions cannot benefit from using such cutting-edge technologies as ChatGPT, among others.
- Potential shifts in the global labor market caused by AI
The labor market changes caused by widespread AI implementation are on the verge of turning into a global threat to many industries, and the museum industry is no exception. The adoption of technologies that could automate and accelerate collection management and artifact handling can potentially reduce the number of museum employees required. The widespread introduction of robots behind the scenes and in the museum halls risks becoming an excuse to lay off the actual museum guides since robots are less prone to making mistakes and easier to maintain, thus more effective. The same trend affects the museum’s experts since adopting and managing massive AI systems for storing and searching for data seems simpler and more resource-effective than keeping and managing human resources.
However, there still remains a list of potential risks and dangers of full digitization and reliance on AI, especially in such a field as the museum. Hallucinations and public caution towards AI and AI-powered tech like robots, caused by such concerns as personal information safety, are one of the most prominent problems that museums should consider while making a decision towards fully relying on AI in their systems.
- Sociocultural effect of AI implementation
Public perception of the museum and its purpose is greatly influenced by the widespread adoption of AI tools in nearly all aspects of everyday life. The perception of AI in cultural institutions, particularly in museums, is evolving, mirroring earlier debates about technology in art. This shift can be analyzed from different perspectives.
Historically, the introduction of technology in art spaces has faced criticism from philosophers and cultural theorists. Figures like Jean Baudrillard and Walter Benjamin expressed caution towards the digitization of art and society, and concerns about authenticity and potential loss of cultural value.
Walter Benjamin, in his classical 1936 essay “The Work of Art in the Age of Mechanical Reproduction,” argued that mechanical reproduction changed its perception and cultural context, albeit democratizing the public access to art. Jeran Baudrillard extended these critiques to the digital age. Baudrillard saw modern art as emblematic of a culture dominated by signs and simulations rather than authentic experiences, as he noted in his magnum opus, “Simulacra and Simulation”, published in 1981.
This raises ultimate questions of the role of technology in modern culture and art, as well as the value of originality in the age of digital reproduction. Indeed, the question of whether a visit to a virtual museum can be equivalent to the physical visit, and exploring an artifact in a virtual space equal to observing it first-hand is an issue every museum has to address before augmenting its exhibitions using digital technology.
Currently, opinions among cultural heritage specialists worldwide are diverse regarding AI in museums. Some experts envision AI-powered innovations as valuable opportunities for education and inspiration. Others express more cautious views, highlighting ethical concerns such as potential bias and privacy issues. These debates are likely to continue as AI technology is still in its early stages of integration into the museum domain.
A survey conducted in June 2023 by UK Heritage Pulse provides insight into the current state of AI adoption in the heritage sector in the United Kingdom. Out of 154 respondents, only 24% were aware of AI being used within their organization, and over 65% had not considered how they might utilize AI in the future.
While this sample size is not large enough to represent definitive industry-wide trends, it does indicate a significant level of uncertainty and limited adoption of AI technologies in cultural institutions. These findings highlight the ongoing process of understanding and integrating AI technologies in cultural heritage institutions, with a mix of enthusiasm, caution, and uncertainty shaping the discourse.
As mentioned before, museums have long been stereotyped as past-oriented institutions focused solely on preserving history and opposing technological progress. This perception has often hindered their digital transformation, causing them to lag behind other cultural sectors open to advancements in digital technology like media art and entertainment. However, this museum paradigm is currently shifting. For example, museums are increasingly recognized as dynamic spaces for contemporary engagement. Digital strategies are becoming integral to museum operations and visitor experiences. Currently, AI and digital technologies are enhancing accessibility and creating interactive, personalized experiences. Recent developments indicate a growing acceptance of AI in cultural institutions: AI is being used to connect cultural institutions with users, improving understanding of visitor preferences. Digital transformation is now seen as crucial for museums’ sustainability and relevance.
The challenges highlighted above underscore the need for careful implementation to ensure AI enhances, rather than undermines, the integrity of cultural institutions. As public perception continues to evolve, museums are likely to play a pivotal role in shaping the cultural landscape of the digital era, balancing technological innovation with their core mission of preservation and education.
Is there a way to address these challenges?
Having discussed the challenges around wide AI adoption by museums, as well as other cultural institutions like archives and libraries, we might have a logical question of whether there is a way of addressing these fundamental issues. Museums all over the world are experiencing the same trend, but respond to it in different ways.
As museums increasingly explore AI technologies to enhance visitor engagement and operational efficiency, European cultural organizations are advocating for structured frameworks to address ethical and technical challenges. In March 2024, the Network of European Museum Organisations (NEMO) and ICOM presented policymakers with actionable recommendations, urging governments to develop a political vision that positions museums as ethical custodians of AI integration. Central to their proposal is the creation of a European AI Innovation Hub – a collaborative platform to share expertise on mitigating risks such as algorithmic bias, AI hallucinations, and potential labor market risks. This initiative aligns with technical safeguards already piloted by institutions like Duke University’s Nasher Museum, which combines human oversight with AI tools to flag inaccuracies, and Terentia’s metadata systems, which use confidence scoring to prioritize curator review.
Meanwhile, workforce strategies, including the Met’s AI literacy workshops and the Rijksmuseum’s cross-departmental task forces, aim to reskill staff rather than replace roles. Critically, these efforts require sustained funding for digitization and infrastructure, as outlined in the European Charter’s call for interoperable, open-source systems that respect copyright and privacy.
When it comes to the response of Korean museums, the National Museum of Korea is at the forefront of digital innovation in the cultural heritage domain. Most notably, ‘The Digital Strategy 2025’, spearheaded by the National Museum of Korea, focuses on transforming museums into digitally immersive spaces through AI-driven initiatives like virtual tours, interactive exhibits, and personalized visitor experiences. Announced in 2021, it presented a 4-year plan to transform museums by integrating them with AI technologies, including robot guides, immersive experiences, and information-providing chatbots. As the National Museum of Korea is considered one of the most technologically advanced museums in the country, it is intriguing to see the results of this strategy, which will most likely be announced to the public in the near future.
Besides, the Korean museum initiatives are not limited to the general issues faced by museums worldwide. Korean cultural institutions explore the wider range of possible applications of AI technology. As an example, The Korea Heritage Agency has adopted AI tools, such as chatbots, to streamline internal processes and improve public access to heritage information. This effort is part of a broader government-funded push to digitize cultural assets, with plans to introduce cloud computing services and expand outreach to marginalized communities. These efforts demonstrate the country’s commitment to leveraging AI for the preservation of its own cultural heritage and improving its public accessibility, all while keeping an eye on the more global obstacles such as data security, bias, and hallucinations.
Overall, throughout our exploration of the phenomenon of AI adoption in museum space, we covered not only its unique history, but also touched upon the issues that currently challenge the widespread AI implementation, and the public opinion on the topic of digital invading the cultural, a concept which can be applied to museums as well. Finally, we outlined some of the possible solutions to the stated problems, discovering that through close institution and government collaboration, even such fundamental problems like data insufficiency and resulting hallucinations, can be tackled with relative ease. That is, only by anchoring AI adoption in transparency and human-centered values, can museums balance innovation with integrity and ensure technologies like LLMs amplify, rather than undermine, their educational missions and, in the nearest future, become a new present for the guardians of the past.
Recommendation for readings
Thiel, Sonja and Bernhardt, Johannes C.. AI in Museums: Reflections, Perspectives and Applications, Bielefeld: transcript Verlag, 2023
Artificial intelligence is becoming an increasingly important topic in the cultural sector. With a decidedly interdisciplinary approach, this book brings together a wide range of critical reflections, practical perspectives and concrete applications of artificial intelligence in museums, and provides an overview of the current state of the debate.
Benjamin, Walter, The Work of Art in the Age of Mechanical Reproduction, Penguin Great Ideas, 2008
Even if not directly tied to AI and museums, this fundamental work by Walter Benjamin explores the implications of digitization of art and culture, which can take many forms, including mass AI adoption. This thought-provoking essay is thought to be one of the earliest works exploring the influence labor automation can have on cultural values of society. The ideas expressed can be applied to museums, as they have always been perceived as the stoic keepers of the archaic, ‘tangible’ culture.
References
- A. Siri, “Emerging Trends and Future Directions in Artificial Intelligence for Museums: A Comprehensive Bibliometric Analysis Based on Scopus (1983-2024),” Geopolitical, Social Security and Freedom Journal, vol. 7, no. 1, p. 2024
- P. JIN-HO, “DIGITAL TRENDS IN THE ERA OF THE DIGITAL MUSEUM,” Public History & Museum, vol. 2, pp. 129–135, 2019
- R. R. Budagyan and Б. Р. Р, “ТЕНДЕНЦИИ ПРИМЕНЕНИЯ ЦИФРОВЫХ ТЕХНОЛОГИЙ В ПРОСТРАНСТВЕ СОВРЕМЕННОГО МУЗЕЯ,” Сфера культуры, vol. 2, no. 1, pp. 61–68, May 2021
- Villaespesa, Elena & French, Ariana. (2019). AI, Visitor Experience, and Museum
- Operations: A Closer Look at the Possible.
- J. Li, W. Wider, Y. Ochiai, and M. A. Fauzi, “A bibliometric analysis of immersive technology in museum exhibitions: exploring user experience,” Frontiers in Virtual Reality, vol. 4, p. 1240562, Sep. 2023
- T.-C. Yu et al., A Systematic Review of Integrating Mixed Reality and Artificial Intelligence in Museums: Enhancing Visitor Experiences and Innovating Exhibit Design. 2025.
- J. Sandahl, “Addressing Societal Responsibilities Through Core Museum Functions and Methods: The Museum Definition, Prospects and Potentials,” Museum International, vol. 71, no. 1–2, pp. iv–vi, Jul. 2019
- Trichopoulos, Georgios & Konstantakis, Markos & Caridakis, George & Katifori, Akrivi & Koukouli, Myrto. (2023). Crafting a Museum Guide Using ChatGPT4. Big Data and Cognitive Computing. 7. 10.3390/bdcc7030148.
- Smithsonian Organization and Audience Research. (2018). An evaluation of pilot Pepper robot program. In Smithsonian Organization and Audience Research.
- Baudrillard, Jean. Simulacra and Simulation. Translated by Sheila Glaser, University of Michigan Press, 1994.
Online article
- 「Artificial intelligence (AI) and the Modern Museum」, https://www.linkedin.com/pulse/artificial-intelligence-ai-modern-museum-bluecadet-interactive/
- 「AI in Museums: What Does Artificial Intelligence Mean for the Industry?」, https://www.terentia.io/thoughts/ai-in-museums
- 「Nemo presents recommendations for AI in museums」, https://www.museumsassociation.org/museums-journal/news/2024/04/european-sector-body-presents-recommendations-for-ai-in-museums/
- 「European museum sector publishes recommendations for AI development in museums to policymakers」, https://charter-alliance.eu/museum-ai-recommendations/
- 「Heritage Pulse: Spotlight on Artificial Intelligence」, https://www.culturehive.co.uk/resources/heritage-pulse-spotlight-on-artificial-intelligence/#fewer
- 「Machines meet museums: Report unpacks AI in heritage sector」,https://techxplore.com/news/2023-11-machines-museums-ai-heritage-sector.html
- 「DIGITAL STRATEGY 2025: Transforming the Museum into a Place for Emotional Healing」,https://webzine.museum.go.kr/eng/sub.html?amIdx=15978
댓글 남기기