How the Public Really Feels About AI in Museums, Libraries, and Archives
Artificial intelligence is no longer some distant notion confined to laboratories or the pages of science fiction. From chatbots that field our queries to algorithms capable of composing entire works of art, AI has quietly – and sometimes dramatically – woven itself into the fabric of everyday life. Yet when it comes to our most cherished cultural spaces – the galleries, libraries, archives, and museums that preserve and interpret human knowledge – how do people actually feel about this technological shift?
A recent study, 'People’s Voice: How Does the Public Perceive the Impact of Generative Artificial Intelligence on Galleries, Libraries, Archives, and Museums', by Kavya Gupta, Jordan Braidwood, and Yuan Li at the University of Alabama, delves into this question with care and precision. Presented at the ALISE 2025 conference, their work explores how ordinary people are talking about generative AI (GenAI) within the so-called GLAM sectors – a collective term for Galleries, Libraries, Archives, and Museums.
Drawing on over 215,000 posts on X (formerly Twitter) from November 2022 to December 2024, the researchers used data science techniques – from topic modelling to machine-learning-based sentiment analysis – to trace how the public discusses AI in cultural life. What they found was a vibrant mix of fascination, creativity, anxiety, and scepticism. In short, people are paying attention, and their reactions vary dramatically depending on which corner of the cultural world is in focus.
'Our findings reveal distinct themes across GLAM institutions and indicate that positive sentiment was most prevalent within museums and galleries and negative sentiment was more frequently linked to archives.'
The report suggests that the conversation around AI and culture is far from uniform. Libraries wrestle with questions of identity and innovation; archives grapple with authenticity and trust; while museums tend to meet AI with curiosity and enthusiasm.
Libraries
If there’s one GLAM sector where AI stirs both optimism and unease, it’s libraries. Posts about ChatGPT automating cataloguing systems, AI assisting with metadata creation, or even writing eBooks have sparked thousands of online conversations. The study noted that the most frequent library-related topic – 'chatgpt ai librarians audiobooks' – appeared in over 8,000 tweets, a clear sign of how tightly AI and librarianship have become linked in the public mind.
For many users, AI offers an opportunity to make libraries more efficient, automate repetitive tasks, and improve access to research tools. Conversations about 'help desk message with Gemini' and 'knowledge sharing' reflected a real excitement about streamlining services. Yet, nestled within that enthusiasm is unease. If AI can answer questions or summarise information, what becomes of the librarian’s human expertise?
'These discussions likely reflect the general public’s concerns about GenAI’s evolving role within libraries.'
This ambivalence runs deep. Some see AI as a digital assistant that could free librarians to focus on the more human aspects of their work – teaching information literacy or guiding deeper research. Others worry it might diminish their professional value, or even replace them entirely. The study’s sentiment analysis showed surprise as the dominant emotion in library discussions – a cocktail of curiosity and apprehension. People seem intrigued by what AI could do but unsure of what it should do.
Librarians, as the paper notes, are already exploring responsible frameworks and ethical codes to handle AI’s integration. The key may lie in collaboration – ensuring AI enhances rather than undermines human skill.
Archives
If libraries are about adaptation, archives are about truth. As custodians of history, archives exist to safeguard authenticity, and many social media users expressed fear that AI could distort or even rewrite it. Phrases such as 'archives AI historical fake' and 'chats deleting archive ChatGPT archived' captured concerns about AI’s capacity to falsify the past.
'Archival users tended to distrust GenAI, assigning lower value to AI-generated archival text compared to human-written content.'
The archival field drew the highest share of negative emotions – notably anger and fear. Some praised AI for improving access to historical records through better search functions and tagging, but others viewed it as a threat to the integrity of history itself. A recurring theme was distrust of AI-generated material, underpinned by ethical unease over accuracy and bias.
This tension between innovation and authenticity raises profound questions. If AI reconstructs a damaged manuscript or fills in missing text, is it preserving history or inventing it? And how can archivists remain transparent about what is machine-generated versus human-authored? The authors argue that answering these questions will be vital to maintaining public trust.
In many respects, archives are where the stakes of AI are sharpest. Unlike art or storytelling, where invention is welcome, archival work demands truth and accountability. GenAI’s arrival forces archivists to reconsider what it means to document reality in a digital world.
Museums and Galleries
In striking contrast, museums and galleries appear to be riding the wave of AI innovation with enthusiasm. Discussions about this sector online radiated excitement and creative energy. The study found that happiness was the most common emotional response, as users shared awe and fascination at AI-driven exhibitions, digital artworks, and interactive experiences.
'Museums and galleries exhibited the highest enthusiasm for GenAI, with the most prevalent expressions of happiness.'
Tweets referencing 'art museum artificial intelligence paintings' and 'gallery chatgpt turn free imagination' captured this wonder. People marvelled at AI’s ability to generate art, analyse style, or act as a personal guide. Tools such as DALL-E, Midjourney, Gemini, and Sora appeared frequently, with users describing them as creative partners rather than threats.
Even so, museum professionals remain aware of bias, authenticity, and copyright. Questions persist about whether AI-generated art should hang beside human creations, and how to credit machine-assisted works. Yet the overall tone is one of optimism. As the study observes, 'Museums and galleries may be more open to GenAI experimentation.' These institutions sit at the crossroads of education, imagination, and storytelling – ideal ground for exploring the creative potential of technology.
In practical terms, AI is already broadening accessibility, generating multilingual guides, and enriching educational content. Some museums use AI to interpret visitor data to design better exhibits; others employ it to generate narratives that bring history vividly to life. The union of art and algorithm is providing fresh ways for cultural institutions to evoke emotion, reflection, and participation.
Emotions Across the Cultural Landscape
One of the most intriguing findings from the study is the emotional diversity across the GLAM ecosystem. Using a refined BERT model, the researchers charted five core emotional responses – happiness, sadness, anger, fear, and surprise – and found each sector had its own distinct emotional signature:
- Museums and galleries: Dominated by happiness and optimism – enthusiasm for creative possibilities.
- Libraries: Defined by surprise and curiosity – open yet cautious about AI’s reach.
- Archives: Marked by anger and fear – concerned about authenticity and accuracy.
'Surprise dominated emotional responses in libraries and archives, likely reflecting curiosity and novelty.'
Sadness and fear cropped up everywhere, often linked to worries about jobs and expertise. Yet beneath that anxiety lies hope: people care deeply about their cultural institutions and want them to thrive responsibly. The data shows that these conversations are not just about technology, but about human values – trust, identity, and belonging.
The study’s charts also highlight the contrast between sectors. Museums had the most positive responses, while archives ranked lowest, reflecting differing attitudes towards risk and innovation. The authors argue that such emotional patterns reveal where dialogue and ethical guidance are most urgently required.
The Road Ahead
Gupta, Braidwood, and Li close their paper with an important reminder: success in applying AI to culture depends not only on what the technology can do, but on how thoughtfully we use it. Every GLAM sector has its own challenges and opportunities, and each must navigate them in its own way.
'Responsible GenAI adoption requires sector-specific strategies, ethical safeguards, and proactive public engagement.'
That engagement starts by listening – exactly what this study set out to do. By turning online chatter into meaningful insight, the researchers have captured a global conversation about how people view AI’s place in culture. The next step is to transform perception into partnership, inviting the public to help shape these technologies.
For libraries, that might mean being transparent about how AI supports rather than replaces staff. For archives, it demands openness, auditing, and careful management of bias. For museums and galleries, it could mean continuing to experiment, but always with authenticity and inclusion in mind.
AI mirrors both the brilliance and the flaws of human nature. It can shed light on the past, enrich the present, and reimagine the future – but only if handled with humility and accountability. Cultural institutions, grounded in both heritage and innovation, are uniquely placed to lead that conversation.
From the quiet corridors of libraries to the bustling halls of contemporary galleries, AI is reshaping how culture is created, curated, and experienced. The public’s reactions – hopeful, hesitant, occasionally fearful – remind us that technology never operates in isolation. It reflects our hopes, biases, and creativity.
Ultimately, this study shows that the debate over AI in GLAM institutions is not one of machines versus people. It is about how technology and humanity can work together to preserve memory, inspire imagination, and widen access to knowledge. If handled wisely, AI might help these institutions become not just repositories of history, but laboratories of the future.
As the authors suggest, the question is not whether AI belongs in our cultural spaces, but how we ensure it belongs well. And perhaps that, more than anything, is the conversation worth continuing.
