Prepared by the researche : ISSAOUI Hana – ISAM SFAX – Gafsa TUNIS
Democratic Arabic Center
Journal of Strategic Studies for disasters and Opportunity Management : Twenty-fifth Issue – March 2025
A Periodical International Journal published by the “Democratic Arab Center” Germany – Berlin
:To download the pdf version of the research papers, please visit the following link
Abstract
Artificial intelligence (AI) is increasingly vital in contemporary society, particularly in cultural contexts, offering unique potential to preserve and enhance heritage assets. AI can transform heritage challenges into opportunities, while also revealing its limitations in this field. Best practices for AI integration across various heritage forms are essential, ensuring all aspects are considered. This vision aims to establish a foundation for understanding AI’s effective use in heritage, be it natural, historical, or cultural. The greatest impact of AI in heritage lies in interdisciplinary collaboration across humanities, sciences, and engineering. AI shapes cultural identity perceptions and can alter how we interact with our environment. Therefore, its adoption can irrevocably change heritage. This study examines AI’s role in heritage, analyzing its meaning for heritage and vice versa. It outlines the study’s structure for reader comprehension. Heritage, both tangible and intangible, is a crucial cultural bond. Its preservation is vital for understanding history and ensuring cultural legacies for future generations. Heritage faces threats from natural disasters and technological development, necessitating effective preservation. Cultural heritage conservation is evolving from a “preserve” to a “sustain” paradigm, viewing heritage as a dynamic system. Technology integration is crucial, but cultural contexts must be considered to avoid unintended consequences. While technology offers opportunities like language translation, it also poses challenges such as fake cultural items. This study explores AI’s integration into heritage, evaluating AI technologies, and emphasizing stakeholder involvement. It acknowledges limitations, such as not fully covering oral heritage, and outlines methodologies, aiming to guide readers towards the research’s wider implications.
1. Introduction
Artificial intelligence (AI) technologies are becoming increasingly prominent and relevant to contemporary society and the significant cultural contexts that shape it. AI has the unique potential to preserve, document, and enhance various heritage assets through a broad and holistic interpretation of what constitutes ‘heritage’ – encompassing everything from places to practices and all the nuances in between. This innovative integration can effectively transform heritage challenges into fruitful AI opportunities, and conversely, it can also identify and address the limitations of AI in heritage contexts. Moreover, it prescribes best practices for the thoughtful and effective integration of AI across the diverse manifestations of heritage, ensuring that all elements are considered. This comprehensive and inclusive vision aims to establish a solid foundation for understanding how AI can be effectively exploited in and for the realm of heritage, regardless of whether the heritage in question is natural, historical, cultural, or belongs to any other category. By fostering a synergy between AI technologies and heritage preservation, we can facilitate a future where heritage is not only preserved but actively enhanced using cutting-edge tools. (Zhang et al., 2023)
This foundation addresses the opportunities, challenges, limitations, and potential pitfalls associated with AI in heritage in nuanced and specific ways. It aims to provide stakeholders, including researchers, conservators, curators, custodians, educators, policy-makers, and technologists, with a means of engaging with AI in the context of heritage, whether currently active in this area or seeking to do so in the future. It also aims to stimulate cross-disciplinary dialogue and knowledge exchange by presenting AI in heritage as a focus of interest to a multitude of disciplines and backgrounds. This is crucial, as the greatest significance and potential impact of AI in heritage lie in interdisciplinary collaboration across the humanities, sciences, and engineering.
The integration of AI into heritage is fundamentally important, as it can shape perceptions of cultural identity and legacy for future generations. AI can also profoundly or subtly transform how the living environment is perceived, understood, and interacted with. Thus, cultural assets and heritage legacies can be irrevocably altered, neglected, and even lost by the adoption and evolution of AI technologies. In addition to addressing the state of AI in heritage, this foundation takes a step back to think critically about what AI means for heritage, and what heritage means for AI. Finally, it outlines the structure of the entire study to provide a roadmap for readership comprehension.
1.1. Background and Significance
Heritage, both tangible and intangible, is the legacy of physical artifacts and intangible attributes inherited from past generations (Zhang et al., 2023). It is a cultural bond that connects human beings across time and space, forming an integral part of human civilization. Preserving heritage is vital for understanding history, appreciating cultures, and ensuring the continuation of historical and cultural legacies for future generations. Throughout history, humanity has faced the challenge of conserving heritage. Natural disasters, man-made calamities, and material decay have threatened the fragile existence of heritage. Despite modern civilization advancements, heritage is more vulnerable than ever due to globalization, urbanization, and rapid technological development. Urgently and critically, heritage needs to be effectively preserved, especially culturally significant heritage.
Cultural heritage conservation practices are rooted in early human civilization. Initially, individual efforts focused on the preservation and perpetuation of cultural items deemed worthy of remembrance. Over time, collective actions emerged, leading to the establishment of formal entities, regulations, and systems for cultural heritage conservation. The “preserve” paradigm dominated cultural heritage conservation for centuries, driven by a good faith belief in the value of heritage. However, the effectiveness of this paradigm is now questioned. With changes in scientific rationality and postmodern critiques of historicism, linear narratives of cultural/temporal progression no longer resonate. There is a return to the “sustain” paradigm, viewing heritage as a socio-natural system for diversity reproduction, not temporally fixed but dynamically co-evolved. The conservation practices exit fixed state preservation procedures, becoming a co-evolution process with heritage meaning negotiations.
Confronted such a paradigm transition, a path forward is to discover methodologies that enable the diverse newly cherished heritage meanings to be reasonably negotiated and the threats to such negotiations effectively mitigated. Integrating technology into the cultural context is a natural progression, enhancing the effectiveness and efficiency of the cultural endeavors. Given its maturity and universality, the integration of widely used technologies into the cultural context is the most straightforward approach, with obvious benefits. However, the specific cultural contexts may not be properly considered, leading to unintended consequences. The technological catastrophes bring challenges: the proliferation of fake cultural items threatens the novelty dissemination; the addiction to time-consuming cultural items endangers cultural diversity; the irreversible technologies convergence traps the cultures in the global dominance of a few technology providers. Nevertheless, technologies also bring opportunities: the loss of language endangers the survival of minority cultures, whilst the technologies’ robustness enables translation across diverse representation systems; the desire for cultural understanding drives the cultural artifacts proliferation. Hence, it is significant to explore methodologies that effectively consider the cultural contexts in technology integration.
1.2. Scope and Objectives
The aim of this study is to explore the integration of artificial intelligence into heritage. The specific objectives to be achieved and the expected outcomes from this study are clearly articulated. The scope encompasses a wide number of issues such as applications, challenges, and future directions of AI within a heritage context. Other key objectives cover the aim of evaluating existing AI technologies and assessing their relevance to cultural assets. In addition, the importance of involving different stakeholders in the discussion and exploration of these topics is underlined. It is acknowledged that it is not possible to cover all these aspects in equal depth, and the intention is mostly to initiate wider exploration of these topics within the heritage sector (Zhang et al., 2023). The methodologies that will be used for these objectives throughout this study are also outlined. Finally, the limitations of the research are acknowledged in order to provide a realistic frame for the findings. For instance, it is acknowledged that while efforts are made to ensure inclusivity in terms of addressing various forms of heritage, some particular aspects are not covered here, such as oral heritage. Other similar limitations are also considered. This section also indicates the significance of the research and guides the reader toward its wider implications.
2. Understanding Heritage
Heritage is foundational to every culture and plays a vital role in the development and sustainability of local communities. It serves as a reservoir of shared memories, experiences, and values that bind individuals together. Heritage encompasses both the physical and intangible manifestations of the rich history of a particular group of people. To better understand its significance, heritage is classified into three primary categories: natural heritage, cultural heritage, and mixed heritage. Cultural heritage, in turn, further categorizes itself into tangible cultural heritage, intangible cultural heritage, and documentary heritage, each representing different aspects of a community’s legacy. Tangible cultural heritage consists of movable assets, such as a wide range of artworks—paintings, sculptures, and crafts—as well as manuscripts that document important historical narratives. Additionally, it includes immovable assets, which are significant structures such as monumental buildings, historical sites, and archaeological locations that have stood the test of time. On the other hand, intangible cultural heritage encompasses a variety of expressions, including oral traditions that are passed down through generations, the performance arts that bring stories to life through music and dance, and social practices that highlight the unique ways communities engage in daily life. It also embraces rituals that mark important life events and traditional crafts that showcase the skills handed down over generations, emphasizing a community’s identity and creativity. Together, these elements weave a complex tapestry that reflects the essence of a people’s cultural identity and their connection to their history.
Heritage, in a broader sense, includes every element of a culture. It encompasses all the traditions, beliefs, languages, monuments, artworks, music, songs, and narratives that were thought to be lost in the era of modernization but have become the defining attributes of communities today. It defines the identity of a culture and a community. Every heritage asset is vital to its community. The intangible cultural heritage assets remind the community of its past, present, and future. They map its journey through time, depicting the changes in thoughts and practices, in turn, scripting its evolution. The tangible cultural heritage assets signify the values and priorities of a community, reflecting its worldviews, moral codes, and social arrangements. They stand testimony to the sensitivity of the community toward the environmental context, emerging technologies, and socio-political frameworks. They are the material embodiments of the community’s aspirations, pride, fears, and anguish, and thus, profoundly affect its collective consciousness.
Heritage assets are at risk. They have been, and continue to be, altered, modified, vandalized, neglected, and destroyed for the sake of “development.” The savage propagation of global capitalism, colonialism, and modernization is wiping out the cultural diversification of the world, imposing uniformity as the world is becoming a “global village.” The natural disasters and political turmoils are further complicating the situation. Many heritage assets are lost forever, while many others are on the verge of extinction. The intangible cultural heritage is particularly vulnerable as the oral traditions and performances pass on the knowledge and skills without a written script. The pristine forms of many heritage assets are destroyed as they are reinterpreted in accordance with the contemporary understanding. The exploitation of a community’s own heritage assets by external forces often leads to an arbitrary interpretation of the culture, resulting in a mockery manipulation of its values.
The current generation is responsible for the heritage assets of the past generations. Provisions should be ensured for the future generations to inherit what the present generations have received from the past generations. This responsibility is further emphasized in the Convention for the Safeguarding of the Intangible Cultural Heritage, which states that cultural heritage are all practices, representations, expressions, knowledge, and skills – together with the instruments, objects, artifacts, and places – that communities, groups, and, in some cases, individuals recognize as part of their cultural heritage. Understood in this broad sense, cultural heritage is not the concern of one particular sector; it ensures the diversity of cultural expressions in the face of homogenizing global forces. Cultural heritage is thus a collective responsibility. The heritage of a community is its collective asset, and thus it is primarily the responsibility of the community to safeguard it. But abiding by the ethical and moral obligations, the outside world should lend a helping hand in the conservation efforts. These underpin the objectives of heritage conservation: it is not merely preserving the objects against deterioration; rather it is ensuring the continuity of the culture (Zhang et al., 2023).
2.1. Definition and Types of Heritage
The term heritage is deeply immersed in a rich and varied array of terminologies that encompass different aspects and interpretations. Usually, in discussions about heritage, the noun is often preceded by a range of adjectives that serve to describe the specific kind or category of heritage being referred to. However, despite this extensive usage of the term in various contexts, the concept of heritage as such is rarely defined in a clear and universally accepted way. In an effort to make sense of this diversity as a foundational starting point for deeper discussions and explorations, heritage can be comprehensively defined as everything from the past that holds significant value in the present and is wished to be preserved and cherished for the future generations to come. Expanding on this definition in a more structured manner, a classification system of the most commonly used types of heritage is introduced, which encompasses several key categories. These categories include natural heritage, cultural heritage, industrial heritage, and intangible heritage. Additionally, a further sub-classification of cultural heritage into tangible cultural heritage and intangible cultural heritage is also commonly utilized and recognized in various discussions and scholarly works. (Zhang et al., 2023) (Ursu et al., 2022)(ROHIM et al.2023)(Ma & Guo, 2024)(Newisar et al., 2024)(Lam et al.2024)(Zhu et al., 2024)
Natural heritage consists of valuable natural environment assets that are critical for the ecosystem. This concept of natural heritage includes unique natural features, significant areas, diverse wildlife, or places that hold both aesthetic and scientific importance in our world. This specific type of heritage primarily focuses on the intrinsic natural value of these assets and emphasizes their critical significance in the broader environmental context. In contrast, cultural heritage encompasses everything that is human-made or has been influenced by human activity. Cultural heritage can encompass a vast array of creations by human beings, ranging from grand statues and intricate sculptures, colorful paintings, impressive buildings and monuments, to historic manuscripts or any other artifacts that, in one way or another, represent the rich tapestry of human culture. Cultural heritage stands as a legacy inherited from our past, which we are tasked with preserving, protecting, and maintaining in the present day, with the hope and intention that it is successfully passed on to future generations. The significance of cultural heritage is profound, particularly within social and cultural contexts. It greatly enriches social and cultural experiences by offering opportunities to delve into diverse ways of life, beliefs, and various viewpoints that shape our understanding of the world. Industrial heritage, a particular subset of cultural heritage, manifests as a reflection of the industrial activities undertaken by humanity and their resulting by-products. Much like cultural heritage, industrial heritage assets can also be tangible or intangible in nature. Typical illustrations of tangible industrial heritage assets include historical factories, vintage industrial machines, and other physical reminders of past industrial activities. On the other hand, industrial work practices, cherished traditions, and specific techniques represent examples of intangible industrial heritage that reflect our shared industrial history. Intangible heritage encompasses non-material assets that hold cultural significance. This type of heritage is something that is passed down through generations, conveyed through teachings, imitation, and dedicated practice. It primarily pertains to the cultural value of assets, while also encompassing other underlying values. For instance, a particular natural heritage asset may carry substantial scientific value in addition to its aesthetic appeal. As a general statement, it can be asserted that a heritage asset possesses multiple values for society, and these values play a crucial role in influencing the dynamic interplay between different categories of heritage types.
2.2. Importance of Preserving Heritage
The world heritage comprises the invaluable legacy inherited from the past, cherished and celebrated in the present, and wisely bestowed for the future. It consists of a diverse array of natural and cultural assets, and a harmonious combination of both forms the classical and widely accepted understanding of the world heritage sites that are of paramount concern to humanity at large. Natural heritage encompasses the aesthetically breathtaking and scientifically significant natural assets that are being diligently safeguarded for the benefit and enjoyment of posterity. Cultural heritage, on the other hand, refers to the rich legacy of the tangible items and their intangible attributes that have been inherited from past generations, preserved and conserved for future generations owing to their profound artistic, cultural, or historic value. The extent of conservation and protection of the world heritage sites on a global scale and particularly in developing nations is increasingly worrisome and requires urgent attention. Thus, the ongoing loss of cultural assets due to neglect, vandalism, and deviance has raised significant concerns for the preservation of our precious heritage. As we move forward, it is essential to recognize the importance of safeguarding these treasures for the benefit of all humanity and to ensure that future generations can appreciate and learn from the diverse tapestry of our collective human experience. (Otero, 2021) (Weiss et al.2022)(Sandu, 2022)(Khan & De Nardi, 2024)
The intelligible reason for preserving heritage is worldwide shared common good held in trust, nevertheless the renaissance of local heritage has prompted manifold reasons to preserve heritage locally. Heritage is time imbued places, practices and artefacts, elaborated on rootedness a sense of belonging and possession and hence crucial for cultural identity and community cohesion. Heritage also tempers change by acting as a point of reference, continuity and stability, and hence is vital in a fast-paced world. Heritage provides a lens to the past, an archive of memory, a source of stories and experience and hence enhances educational opportunity and superannuates numerous creative endeavours. Further, heritage tourism is a growing sector in global tourism industry and therefore preservation of heritage is an economic opportunity, vital for local economies.
To many, heritage is an embodiment of the past in the present and therefore a sense of connection to history, wholesome objects that renders social well-being and environ to be at ease and hence there is an emotive draw towards heritage. On the contrary many cultural assets are in the process of deviance, neglect and oblivion contested by socio-political changes, often jeopardizing the communal identity they once forged. Despite its alleged parochialism, world over and particularly in developing nations, historic assets under the poverty alleviation discourse are considered deviant, unworthy of investment, and cultural assets are the first to erode. Therefore, the urgency to act if cultural assets ought to be preserved beyond a burgeoning lucre. With the emergence of nascent culture based initiatives, it is argued that the conservation needs to be integrated, bottom-up, and participative with the community at its core, otherwise the sense of ownership would be lost deliberating the cultural asset deviant. Nevertheless the contemporary conservation discourse is plagued with inadequacies, thus necessitating to re-think and re-frame the conservation paradigms.
Heritage is considered a collective memory, an inalienable common good, and therefore its appropriation must be shared in trust through co-operation and collaboration (Skrzypaszek, 2012). With the notion of shared good, it is argued that post-colonialism appropriation of heritage in terms of power has rendered the historic assets of the developing world inadequately treated and trivialized, thus culturally deviant. Cultures worldwide undergo change, broadly external perturbation either by conquest or globalization, in response to perturbation some elements of culture tend to stratify, whilst some continue to function as before or metamorphose. Globalization upfront homogenizes culture, nevertheless local heritage articulates a counter response, a renaissance, craving for cultural distinctiveness and therefore an emblem to local identity community embedding.
3. Artificial Intelligence Fundamentals
Beginning in the 1950s, artificial intelligence (AI) emerged as a groundbreaking transformative technology that mimics human reasoning and learning, profoundly affecting many aspects of society and reshaping various industries. AI embodies the concept of machines that possess the capability to perform tasks that would traditionally require the type of intelligence that humans exhibit, such as understanding natural language, reasoning, problem-solving, or even engaging in complex games. Machine learning (ML) plays a crucial role in this field by enabling machines to learn from vast amounts of data and continually improve their performance over time as they gain experience. Natural language processing (NLP) empowers machines to not only understand but also generate coherent human language, effectively bridging the communication gap between humans and machines. Furthermore, computer vision (CV) endows machines with the ability to “see” and interpret the visual world around them, enabling the analysis and understanding of images or videos, which is essential for applications ranging from facial recognition to autonomous driving. These advancements in AI have the potential to revolutionize various sectors, including healthcare, finance, and entertainment, leading to a future where machines assist humans in unprecedented ways. (Colavizza et al., 2021)
AI applications can be classified as narrow or general. Narrow AI applications are goal-oriented and focus on specific tasks like language translation, web searches, and personal assistants. Conversely, general AI applications possess the intelligence seen in humans and can perform various tasks. Generally, AI applications can be classified as perception, reasoning, and action. Heritage applications primarily use AI perception technologies to enhance reasoning and action capabilities. The power of AI lies in its ability to process large amounts of data quickly and effectively, recognize patterns in data, and automate or augment tasks. These capabilities enable machines to perform human tasks more effectively. AI technologies have the potential to significantly impact decision-making processes by automatically making decisions based on learned models or augmenting human decision-making. Forty years after the first debate on the integration between natural and cultural heritage, AI technologies have the potential to radically change how heritage is understood, preserved, and disseminated.
3.1. Overview of AI Technologies
As artificial intelligence (AI) technologies mature, they are gradually being applied to various fields. Some uses are straightforward, such as having automation cameras on public transport to monitor and ticket fare evaders. Others are more sophisticated, requiring refined AI analysis of large data sets that would be impossible for humans to manage alone. For instance, credit rating companies have an elaborate system of algorithms and neural networks that analyze 6,500 attributes for every company in their database, allowing them to make precise ratings. On a simpler level, social media platforms assess content by auditors and algorithms to identify hate speech, misinformation, or other violations (Cetinic & She, 2021). While some AI technologies have flaws, such as the misidentification of people by surveillance systems, most are precise and practically relevant. In industries where auditing or analysis of complex data sets is crucial, AI is becoming common. For example, some music streaming services use AI technologies to audit their vast libraries to identify potential copyright violations. The pharmaceutical industry uses AI systems to assess the efficacy of drug compounds, while the banking industry employs them to identify money laundering activities. Real estate agencies use AI to analyze building price trends, and the insurance industry to review fraud attempts. In short, current and near-future implementations of AI technologies are concerned with high-precision analysis of large data sets.
While AI advancements are impressive, the technologies still have limitations. Most AI technologies are not widely accessible due to the resources and expertise needed to develop them. Even accessibility brings challenges. For example, some painting restoration AI technologies are online, but results often require expert understanding and insight for meaningful output. Therefore, some implementations provide either impressive results or meaningless output that a layperson cannot interpret. Additionally, some developers are not motivated to make results easily interpretable. A recent socio-political experiment where text-to-image AI systems were exposed to artwork resulted in racist outrage illustrations. While some of these systems can generate impressive images when fed with text and art style, the developers did not include use instructions, limiting the technology to those willing to experiment for hours.
In contrast, other relevant systems are designed to be easily accessible and produce understandable output, as they were developed alongside research in art analysis and interpretation. Considering the current trajectory of AI technologies, a short overview to assess relevance and applicability in heritage context is provided. Of course, it is impossible to cover everything. Most text-to-image AI technology applications are trivial reproductions or parodies of already existing artworks rather than an attempt to analyze and create art. Still, recent research on how these technologies understand art illuminates relevant aspects of AI application in heritage.
3.2. Applications of AI in Various Fields
Artificial Intelligence (AI) is a term encompassing various approaches, methods, and applications. It can be broadly defined as a technology that simulates human intelligence through either learning from experiences or data. AI is seen as a decisive force for future development, with various initiatives, efforts, and examples emerging from industry, academia, and civil society. The impact of AI on culture and cultural sectors is still being discussed. Nevertheless, strategies addressing the impact and applications of AI on culture and cultural heritage are coming to existence. The most concrete examples often stem from the industrial sector. AI applications pervade daily life and are continuously being developed and adapted to new domains and considerations. Prior to looking at sector-specific examples of AI applications, the breadth of AI applications in various disciplines and fields is discussed.
AI has been successfully applied to various fields, including population health (Lavigne et al., 2019), health care (Klumpp et al., 2021), finance, education, and traffic. Each application often has common characteristics, challenges, or opportunities for those considering implementing AI. Notable examples of AI applications in some of these fields are illustrated to inspire innovative thinking regarding what can be done similarly in the context of, or addressing the needs of, heritage and heritage conservation. AI can, and likely will, enhance efficiency in many actions, tasks, or processes. AI can be trained to perform these tasks similarly to human experts but usually much faster. In addition to solely enhancing efficiency, AI can also augment and improve analytical capabilities. Simple or more complex data models trained with AI can make inferences and provide analyses that humans cannot achieve, even with a comprehensive understanding of the modelled phenomena. For example, predictions can be achieved using large amounts of data that would be impossible for humans to process manually.
The significance and effort put into developing various AI applications in these fields already drive the discussion regarding possibilities and implications in the cultural sectors or focusing on heritage. Different, and often more challenging, considerations face adaptive applications in different contexts or sectors. Still, the groundwork for linking AI advancements with prospects in aspects of culture and heritage concerns is laid out, as the breadth of AI applications is presented. It can be shown how AI applications pervade various fields, including those similar to the considerations of, and approaches taken within, a cultural focus. However, addressing fields currently outside the prospects and applications of AI is equally important. Each field has challenges that need to be addressed before implementing AI applications, and the considerations focusing on culture and heritage often confront similar challenges, albeit within different contexts.
4. AI in Heritage Conservation
Artificial intelligence (AI) is widely recognized as the most rapidly developing frontier of digital technology today. This remarkable field is currently exploring its immense potential to not only boost innovation in various sectors but also to facilitate sustainable growth and address the significant challenges posed by aging populations around the world during this transformative new decade. Between the years 2021 and 2023, a diverse array of selected or completed artificial intelligence applications in various leading cities globally will be showcased, highlighting their impact and importance. On another note, heritage conservation entails a comprehensive set of principles and practices geared towards the preservation, protection, restoration, and maintenance of our rich cultural heritage. This field encompasses both tangible heritage—such as invaluable artifacts, monuments, and historical sites—as well as intangible heritage that includes cultural traditions, rituals, and the many languages that exist.
In recent years, a variety of artificial intelligence (AI) technologies have been actively explored, tested, and developed for truly innovative digital preservation and protection of heritage resources across various domains. Alongside this, significant strides have been made in enhancing public engagement and education through the use of cutting-edge virtual and augmented reality technologies. This text offers a detailed review of the methodologies employed, which include automated data collection and intricate data mining processes, as well as comprehensive analysis of diverse heritage resources. Additionally, it presents a range of innovative applications that have been undertaken, alongside specific case examples that showcase successful integration of AI within the realm of heritage conservation. The development of AI technologies has the potential to open up new pathways for revitalizing our cultural heritage and resources, especially in relation to addressing urgent global challenges such as climate change, the impact of the pandemic, and the issues stemming from an aging population. It is crucial to underscore that effectively realizing the full potential of AI in the field of heritage conservation requires collaborative interdisciplinary efforts from various sectors. Furthermore, an exploration is carried out on the cultural and creative products (CCPs) produced by AI New Year Prints, along with an analysis of the feasibility of utilizing AI-generated artistic creations to support and enable the sustainable development of intangible cultural heritage throughout the globe. (Zhang et al., 2023)
Archives hold a record of individuals and institutions, documenting their actions and activities for the purposes of accountability, compliance, and memory. They have traditionally been a privileged resource for researchers and scholars engaged in the reconstruction of past events. The convergence of AI and archival science and practices has the potential to significantly impact the ways archives are created, organised, accessed, and used. Awareness of the challenges posed by the integration of AI technologies in recordkeeping and archival practices is growing across the archival community (Colavizza et al., 2021). A survey of the growing explorations of AI technologies for archives is provided, focusing on approaches that have already had a significant influence on existing recordkeeping practices. Four distinct areas of integration and emerging trends are identified. Interdisciplinary exchanges between archivists, AI researchers, and computer scientists will be necessary to ensure that AI applications in archival practices are coherent with the theoretical foundations of the discipline.
4.1. Digital Preservation Techniques
Focusing on the digital techniques that have been devised for the preservation of heritage assets, with an emphasis on the application of artificial intelligence (AI) technology, several digital activities that are currently performed by human operators will be discussed. The range of activities runs from the 3D scanning of artifacts to the digitization of documents, from the virtual archiving of processes to the use of technology for monitoring and analysis, and so on (Ramos, 2018). All of these activities aim to produce digital representation data of heritage objects and may be considered digital preservation techniques. The digital representation can enhance the accessibility of heritage objects, both digital born and physical, while the physical objects are conserved, either actively or passively.
Most techniques now in use are labour-intensive, with many processes relying heavily on the manual efforts of operators. As digital technology has become sufficiently mature, there is a good opportunity to use AI to automate certain processes and streamline the technique as a whole. AI is already in use for some basic applications, although many possibilities are still available with regard to the future application of AI in conservation. Metadata and documentation are the key factors of digital preservation data that ensure the long-term preservation and utility of the data. The digital environment, however, presents many challenges with regard to data storage and the accessibility and retrievability of data. These challenges apply equally to digital-born data and data that have been digitally preserved. Some successful cases from the real world illustrate the application of digital preservation techniques. These cases are mostly from research institutes or museums, where certain digital preservation techniques have been implemented either to heritage objects or heritage processes (Colavizza et al., 2021).
4.2. Virtual and Augmented Reality Applications
Virtual and augmented reality technologies have been adopted to various extents in heritage contexts. Generally, this kind of technology creates immersive experiences for the audiences and brings them closer to the cultural assets (Bekele et al., 2018). In recent years, physical artifacts have been closely bridged with digital experiences through the efforts of many artists, researchers, and institutions. Instead of the traditional practice of simply exhibiting the heritage, with the integration of virtual and augmented reality technology, the efforts on heritage now focus on constructing an environment that provides the audiences not only an observation chance but also an immersive experience that promotes their engagement with the heritage. Several successful projects of applying VR and AR in heritage are investigated (Petridis et al., 2013). Although the sophistication of the interaction in these projects varies, they nonetheless open up a new realm of exhibiting heritage. These technologies could bring the public closer to the heritage and greatly increase the accessibility of the heritage. It is also noted that, unlike most of the contemporary practices on heritage, these projects prefer to use simple and even crude 3D models instead of sophisticated high-quality ones. This choice highlights that the focus is put on how to use the technology instead of how to create the most visually appealing exhibits.
In addition to engaging the audiences with the heritage, these virtual and augmented reality applications hold great potentials in deeply educating the audiences about the heritage. For example, the 3D virtual exhibition in the Herbert art gallery allows the audiences to continuously delve into a complicated historical narrative of a painting, which could hardly be conveyed by text-only catalogues or even by on-site observations. The technology not only provides an engaging experience but also enables imparting the knowledge more accurately than the traditional means. The educational capability makes this technology particularly appealing to the institutions of heritage, history, art, and similar fields. Apart from education, this technology is also considered a vital means in the marketing of the heritage by creating eye-catching immersive experiences and raising the curiosity of the audiences towards the traditional experiences. Nevertheless, the adoption of this technology is not always smooth. Like many other emerging technologies, the virtual reality applications for the Herbert museum exhibit have experimental status until the technology becomes prevalent in the daily life of the public. Concerns on how this technology would affect the traditional experiences are also raised. Despite these challenges, the consideration of applying virtual reality in heritage illustrates an innovative way of engaging the public with the heritage and enriching the contemporary understandings of it.
4.3. Data Mining and Analysis
With the gradual growth of data related to cultural assets, the extraction of meaningful patterns has become crucial to their understanding and preservation. In simple terms, data mining involves searching for trends, patterns, or anomalies in sets of data that are otherwise not visible, while analysis refers to a complex set of operations conducted on that data to transform it into information (Otero, 2021). For heritage professionals, these mining and analytical techniques can produce knowledge that assists in understanding the past and present context of heritage assets, including the political, social, economical, environmental, and even scientific conditions in which they were created. Furthermore, this knowledge can help identify trends in data, clarifying the cause and time of certain events, and supporting future projections that inform decision-making. The advancement of Artificial Intelligence (AI) technologies is allowing for the more straightforward implementation of complex data mining and analysis methods, and the generation of insights that would be impossible or extremely time-consuming to derive manually. There is therefore a strong argument that the application of these techniques can considerably assist heritage professionals in developing better-informed strategies for the identification, preservation, and intervention of heritage assets.
Currently, most data related to a particular area of heritage interest exist and are readily accessible in publicly available databases. However, the exploitation of this data to extract useful information has not been conducted, and only simple mining and analytical procedures have been performed. On the other hand, the implicit ethical implications regarding the use of this data must be considered, particularly concerning ownership of cultural information (Abgaz et al., 2021). Although data mining and analysis can reveal crucial insights concerning cultural understanding, the potential for misuse also exists. As such, when used outside its originating community, the extracted knowledge could exploit cultural information without proper credit, recognition, or compensation. With the exponential growth of publicly available information and data mining capabilities, similar situations have arisen for other cultural situations, leading to increasingly challenging disputes concerning ownership. These issues must be carefully considered before any data mining exercise is undertaken. Three examples of successful data mining applications in a heritage context are provided, followed by a discussion of general challenges that arise when considering the integrity and accuracy of the input data.
5. Challenges and Ethical Considerations
The development and application of artificial intelligence technology on cultural heritage brings not only opportunities and hope but also challenges and ethical considerations. In recent years, there have been many cases using AI technology to analyse and treat cultural heritage. However, there is still a lack of discussion regarding the ethical implications and challenges that may arise from the use of such technology. As an emerging technology, AI raises many concerns about privacy and security. Heritage is sensitive data. The development, application, and use of AI technology on heritage data are highly related to privacy and security issues (Radanliev et al., 2024). There is a need for systematic analysis and discussion regarding issues and challenges tied to privacy and security when integrating AI technology into heritage. Research on AI technology in the heritage context should put the privacy and security issues and challenges at the most priority level, with the hope that it could bring awareness on the importance of privacy and security and promote the discussion on a clearer governance framework in addressing such issues. In addition to privacy and security issues, there are broader discussions on the ethics of AI technology, particularly regarding fairness and bias. In particular, it examines the potential bias issues in AI algorithms in jeopardizing the integrity of heritage data. It is believed that AI applications on heritage should be under continuous scrutiny, and accountability mechanisms should be introduced in such a way that the outcome of AI applications will be transparent and fair. As AI technology becomes increasingly prevalent in everyday life, it is critical to begin discussions on how to prevent biased outcomes from happening in the heritage context. Potential strategies are proposed in tackling bias issues in AI applications relative to heritage. As the awareness of the potential bias issues in AI technology widely used in various aspects grows, it is highly suggested to start adopting ethical guidelines in designing and developing AI technology, particularly in the heritage context. Many organisations and groups have started to propose ethical principles in deploying AI technology. It is believed that discussions should be encouraged at the international level to ensure that AI technology, when applied to heritage, will abide by such ethical principles. AI technology has the potential to bring significant benefits to cultural heritage, but some pressing issues regarding privacy, security, and fairness need to be addressed to ensure the responsible deployment of this technology.
5.1. Data Privacy and Security
Data privacy and security concerns take on a primary role within heritage contexts, which involve the sensitive handling of place-based information about cultural assets, collections, and communities. Unchecked data freedom could increase the risk of cultural dispossession and loss, the misuse of cultural narratives, or the obsolescence of languages and heritage systems (Murdoch, 2021). The importance of a care-based and precautionary approach regarding data handling in heritage is thereby immediately asserted. With the broad uptake of AI systems, there is a particular need to be transparent about the implications of not handling data responsibly. This includes an outline of possible ethical, legal, and cultural liabilities for an unjustifiable or careless sharing of data. Although one might anticipate the loss of control over heritage data, it is instead argued that enactments of AI systems risk being unreasonably damaging to data subjects if the significance of the data being enacted is not sufficiently acknowledged. Robust protection measures must be in place to maintain data confidentiality and collective trust in what is publicly shared (Korobenko et al., 2024). There is currently no guarantee that cultural data acted upon through AI systems will be free from risk. AI-induced obsolescence, misrepresentation or commodification of cultural narratives, and unwanted disclosures or predictions are relevant risks. In the cultural sector, there have been high-profile instances of data breaches that have resulted in unwanted disclosures of sensitive information. To avoid the replication of similar risks in heritage contexts, best practices in data security are therefore needed. This section discusses relevant frameworks and regulatory standards that can help stakeholders ensure that data security and protection by design is put into practice. When data protection measures are not put in place, liability for damage enacted through cultural data would fall entirely on data subjects. Specific awareness-raising discussions take place with data guardians in situations where community consultation is needed to make sure that data custodians do not have the right to control sensitive information. Such discussions explicitly address the role of community engagement in ensuring the integrity of the data being shared. The aim is for stakeholders to become aware of the need to be proactive in implementing the necessary security arrangements, rather than relying only on enactments being innocuous as they are presently designed. Specific duties associated with a mutuality of trust and the possible dissemination of racist narratives make it inappropriate to share cultural data without sufficient accountability in the data security mechanisms being employed.
5.2. Bias and Fairness in AI Algorithms
Bias and fairness in Artificial Intelligence (AI) algorithms is a critical concern for heritage and other fields. AI algorithms can output unwanted bias due to various reasons, like flawed training data, and the design and implementation of algorithms (Leavy et al., 2020). In the context of heritage, biased AI may lead to misrepresentation, under-representation, or non-representation of heritage, places, and communities. Heritage narratives that go against the dominant socio-political narratives may be erased, leading to homogenisation and marginalisation of heritage. For these reasons, addressing bias concerns becomes fundamental in any AI application to the heritage field. The importance of equitability in the development of AI is stressed. Without equitable practices in the design and development of AI algorithms, claims of inclusiveness from developers and cultural institutions are hard to accept. Cultural narratives are often represented inclusively via dominant technologies that do not allow for the marginalised representation of heritage. Even when these technologies intend to widen inclusiveness, they often trigger exclusion in representation (Zajko, 2020). Strategies to mitigate bias issues in AI models are proposed. These include conducting fairness assessments, retraining models to limit bias, and training AI models with diverse datasets. It also emphasises that operational stakeholders have a primary role in ensuring responsible usage of AI practices. These stakeholders are responsible for upholding ethical standards and should not shift this responsibility to developers. It is asserted that those applying AI in heritage contexts should fully acknowledge the dangers of AI and act to eliminate them. By recognising and addressing these concerns, it is possible to develop AI solutions that are fair and unbiased for heritage. Ultimately, it is argued that integrating ethical considerations is necessary throughout the AI development lifecycle.
6. Case Studies
A series of case studies is presented, showcasing the effective integration of AI technologies in the field of heritage. These projects illustrate how AI has been used to address specific practical challenges in heritage, spanning a wide range of applications, from digital archiving to predictive analytics. This section examines a selection of case studies in greater depth. Each project has successfully integrated AI technologies into heritage and includes not only successful outcomes but also the challenges that were faced, highlighting valuable lessons learned along the way. The diversity of these case studies also demonstrates that AI-based heritage solutions can take various forms, regardless of differences in heritage context. In addition, the ethical issues that each project sought to address are examined, emphasizing that ethical considerations can and should be embedded in the development of AI applications. Collectively, these case studies provide inspiration and best practices for future AI heritage projects and add a practical dimension to the theoretical discussions of earlier sections.
The selected case studies are as follows: the Archives and AI collaboration between the Netherlands National Archive and several Dutch research universities; a collaboration between the University of South Carolina and Historic Columbia to create a deep learning model to transcribe handwritten text in a historic digital archive; the use of Convolutional Neural Networks by the City of Florence to classify and manage the risk of heritage trees; and Deep Learning for Digital Preservation, a project by the University of Iceland to automatically detect and transcribe Icelandic runes in web videos. Each of these case studies highlights how AI has been used to address a public heritage need. While the projects differ in size, budget, and outcomes, they share a commitment to responsible AI heritage applications. Each project undertook efforts to ensure ethical applications of AI within the proposed systems, grounded in established ethical frameworks or guidelines in the heritage or AI fields (Colavizza et al., 2021).
6.1. AI Projects in Heritage Conservation
Within the heritage sector, a collection of noteworthy Artificial Intelligence (AI) projects have been undertaken to address the challenges of heritage preservation. These projects stand as illustrative examples, showcasing diverse methodologies and technologies employed for the effective conservation and engagement of tangible and intangible heritage. They encompass various aspects of heritage, including architecture, oral traditions, natural heritage, handicrafts, and related social activities. Through these accounts, the contributions made through the collaboration among interdisciplinary project teams are highlighted, underscoring the value of combining different expertise. In doing so, the successes and challenges encountered in implementing these projects are explored, aiming to provide a comprehensive overview of efforts made in this emergent field (Zhang et al., 2023). Each project is presented as a self-contained narrative, explaining the context, specific AI technologies used, outcomes achieved, and lessons learned. While some deliberation over theoretical frameworks is offered, the main focus is on practical applications and real-world implications. By showcasing the current state-of-the-art in AI projects applied to heritage, it is hoped that this collection of examples will stimulate interest and inspire stakeholders from various fields to consider how AI technologies could be harnessed for the transformation of heritage (Palma, 2019).
Most of the presented projects arose from a desire to contain or halt the loss of heritage that is considered culturally relevant to some communities. Many are driven by a sense of urgency to safeguard heritage resources that are threatened by globalization and homogenization. Others are motivated by the possibility of new technologies that have the potential to positively transform heritage’s outreach, usability, and understanding. Still, others emerged from the conviction that even the most pioneering technologies should ultimately serve humanity and promote cultural diversity. Attempts to implement AI technologies in the heritage field have sometimes been initiated by scientists from technological backgrounds seeking application fields for their developments. In other cases, attempts have been made to introduce emerging technologies to scientists and practitioners in the humanities and social sciences. It is essential to note that while some projects were initially met with skepticism, a passive or critical stance toward emerging technologies soon transformed into enthusiastic engagement. The articulation of techno-optimistic visions of the future often significantly shaped the initial encounters between AI and heritage. The impact of these projects on local communities and cultural perceptions and understandings of the relevance of heritage is also questioned. As “local” projects strongly attract “outsiders,” the motivations behind such interventions are interrogated.
6.2. Future Directions
Artificial intelligence technologies applied to heritage are still in their infancy. Nonetheless, this text strives to justify the significance of AI technologies by highlighting several trajectories. First, there are many emerging technologies yet to be fully explored. Text analysis via machine learning is a largely untapped avenue for investigating historical texts, as are generative algorithms that may replicate traditional craftsmanship and create socio-culturally relevant designs. Scanners providing RGB, depth, and thermal images are all becoming more affordable, meaning that experimenting with several data types could further improve analyses. These technologies could be particularly beneficial in times of changing climate, since they assist heritage stakeholders in keeping up with rapid changes by developing smart, learning processes. So far, the implementation of technologies has primarily been a one-off investment; continuous adaptations to changing needs would be prudent (Colavizza et al., 2021). A more coherent strategy for technology is required that combines the safeguarding technology with emerging technologies. Moreover, technology and cultural sectors need to work together as much as possible. Experts contemplating technological applications in culture tend to be found in an academic context rather than in tech companies, meaning that research will drive many future initiatives. Still, heritage should be a priority for some technology companies focused on societal challenges. Although some attention has been paid to the risks and challenges that new technologies might bring, these are not insurmountable. Transparent partnerships using data that cannot be misused for other purposes could mitigate data ownership issues. Furthermore, having numerous institutes with similar concerns prevents a few from dominating the discussion. Finally, ongoing research and experiments are needed to discover how AI could be most effectively applied in heritage. The use of AI in the broadest sense—spanning from automating simple tasks traditionally undertaken by humans to machines undertaking tasks previously impossible without human intervention—could be perceived as an opportunity and a threat. Some processes can and will be automated, but creative decisions will remain in human hands, especially with regards to questions of value. Ultimately, the future envisaged is one in which heritage remains vibrant through mutually beneficial collaborations between technology and culture.
7.Emerging Technologies
As technology advances, new opportunities to enhance heritage conservation are unlocked. Emerging technologies with great potential to be integrated within the heritage sector are defined. Generative design, responsible acquisition and smart use of data, and advanced AI methods to create immersive experiences and engage users at different levels are specific attention points throughout the dialogue. Generative design is discussed as a technology with potential in the heritage and conservation sector; an example is a proof of concept using generative design to create unique 3D printed replicas of broken archaeological artifacts to provide new insights in their analysis and study. Methods used for designing three replicas of different styles, fractures, and analysis depths are shared. The influence evolving technologies have on user engagement, participation, and experiential learning is also discussed. Experiments that actively involve visitors in the creation of heritage using technology, such as sand molds 3D printing, generative modeling, and computer numerical control milling, are shared. Emerging technologies are tested and demonstrated in real-world settings; lessons learned are shared, including difficulties and considerations. Likewise, the importance of finding a balance between preserving cultural integrity and driving technological advancement by adopting a responsible and ethical approach in using technology is emphasized. The discussion highlights the importance of finding common ground between disciplines and encouraging cross-pollination opportunities to foster innovation and create outside-the-box solutions. Examples of emerging technologies are tested and demonstrated in real-world scenarios, bringing attention to important aspects learned and considerations that should be taken into account when integrating them into the heritage context. Ultimately, a hopeful picture of the future is presented where technology is carefully and responsibly integrated within heritage, enhancing and expanding what is possible (Palma, 2019) (Zhang et al., 2023).
8. Conclusion and Recommendations
This study examined the integration of Artificial Intelligence (AI) into heritage, considering various aspects such as ethical implications, accountability, job displacement concerns, accessibility challenges, and cultural sensitivity issues. Through a literature review, the current state of AI applications in the heritage sector was mapped. It was observed that while technological advancements are being explored, the need for cooperation and shared understanding between technologists and heritage professionals is crucial. The importance of addressing ethical implications throughout the research process was emphasized, highlighting the potential of AI technologies as innovative and transformative solutions for the heritage sector.
Recommendations were provided to ensure the responsible application of AI within heritage. Best practices for engagement between heritage and technology sectors were suggested, emphasizing accountability, ethics, transparency, and inclusivity. Engaging communities affected by AI applications from the beginning was highlighted as essential for ensuring purpose alignment and sensitivity to potential biases. Continuous dialogue between technologists and heritage professionals was encouraged to enhance understanding and knowledge-sharing, which is vital for the successful application of AI technologies. Additionally, ongoing research is needed to explore how AI can contribute to overcoming future challenges while fostering a culture of learning and adaptation within the heritage sector.
Finally, a framework was proposed to address the ethical implications discussed earlier, focusing on responsible use of AI technologies in heritage. The aim was to consolidate key findings and contributions of the study, providing a starting point for future discussions and considerations. By encouraging stakeholders to actively engage with the proposed framework and consider new questions, the hope is to promote a joint journey into the transformative potential of AI technologies for heritage.
References:
Abgaz, Y., Rocha Souza, R., Methuku, J., Koch, G., & Dorn, A. (2021). A Methodology for Semantic Enrichment of Cultural Heritage Images Using Artificial Intelligence Technologies.
Bekele, M., Pierdocca, R., Frontoni, E., Malinverni, E., & Gain, J. (2018). A Survey of Augmented, Mixed and Virtual Reality for Cultural Heritage.
Cetinic, E. & She, J. (2021). Understanding and Creating Art with AI: Review and Outlook.
Colavizza, G., Blanke, T., Jeurgens, C., & Noordegraaf, J. (2021). Archives and AI: An Overview of Current Debates and Future Perspectives.
Khan, M. S. & De Nardi, S. (2024). The affectual-social ecology of cultural artefacts: Illegal markets and religious Vandalism in Swat Valley, Pakistan. Forum for Social Economics.
Klumpp, M., Hintze, M., Immonen, M., Ródenas-Rigla, F., Pilati, F., Aparicio-Martínez, F., Çelebi, D., Liebig, T., Jirstrand, M., Urbann, O., Hedman, M., A. Lipponen, J., Bicciato, S., Radan, A. P., Valdivieso, B., Thronicke, W., Gunopulos, D., & Delgado-Gonzalo, R. (2021). Artificial Intelligence for Hospital Health Care: Application Cases and Answers to Challenges in European Hospitals.
Korobenko, D., Nikiforova, A., & Sharma, R. (2024). Towards a Privacy and Security-Aware Framework for Ethical AI: Guiding the Development and Assessment of AI Systems.
Lam, J. M., Makhbul, Z. K. M., Aziz, N. A., & Ahmat, M. A. H. (2024). Incorporating multidimensional images into cultural heritage destination: does it help to explain and analyse better?. Journal of Cultural Heritage Management and Sustainable Development, 14(4), 563-580.
Lavigne, M., Mussa, F., I. Creatore, M., J. Hoffman, S., & L. Buckeridge, D. (2019). A population health perspective on artificial intelligence.
Leavy, S., O’Sullivan, B., & Siapera, E. (2020). Data, Power and Bias in Artificial Intelligence.
Ma, Z. & Guo, Y. (2024). Leveraging intangible cultural heritage resources for advancing China’s knowledge-based economy. Journal of the knowledge economy.
Murdoch, B. (2021). Privacy and artificial intelligence: challenges for protecting health information in a new era.
Newisar, M., Selim, G., & Li, M. (2024). Place-based perspectives on understanding the value of sustainable heritage-inspired arts and crafts in Jordan. Sustainability.
Otero, J. (2021). Heritage Conservation Future: Where We Stand, Challenges Ahead, and a Paradigm Shift.
Palma, V. (2019). Towards deep learning for architecture: a monument recognition mobile app.
Petridis, P., Dunwell, I., Liarokapis, F., Constantinou, G., Arnab, S., De Freitas, S., & Hendrix, M. (2013). The Herbert Virtual Museum.
Radanliev, P., Santos, O., Brandon-Jones, A., & Joinson, A. (2024). Ethics and responsible AI deployment.
Ramos, H. (2018). Mixed Reality for Historic Preservation.
ROHIM, R., SAPTAWURYANDARI, N., FUAD, K., SUPRIADI, A., SAEFUDDIN, S., & YULIANTO, A. (2023). Reyog Pomorogo Monument and Museum To Promote Historical Literacy and Cultural Tourism. Quantum Journal of Social Sciences and Humanities, 4(5), 214-222.
Sandu, I. (2022). Modern aspects regarding the conservation of cultural heritage artifacts. International Journal of Conservation Science.
Skrzypaszek, J. (2012). Intangible Heritage and its Role in the Formation of Social and Personal Identity.
Ursu, V., Chiriac, N., & Bostan, I. (2022). Heritage—Polysemantic Analysis of the Concept: Criteria, Forms, Definitions, and Particularities. Open Access Library Journal.
Weiss, T. G., Bokova, I., Adams, S., Al-Sabouni, M., Appiah, K. A., Assomo, L. E., & Bandarin, F. (2022). Cultural heritage and mass atrocities. Getty Publications.
Zajko, M. (2020). Conservative AI and social inequality: Conceptualizing alternatives to bias through social theory.
Zhang, B., Cheng, P., Deng, L., Hanim Romainoor, N., Han, J., Luo, G., & Gao, T. (2023). Can AI-generated art stimulate the sustainability of intangible cultural heritage? A quantitative research on cultural and creative products of New Year Prints generated by AI.
Zhu, X., Oberman, T., & Aletta, F. (2024). Defining acoustical heritage: A qualitative approach based on expert interviews. Applied Acoustics.
