Public History has always concerned the relationship between historical knowledge and wider society. Museums, monuments, archives, exhibitions, documentaries, oral histories, heritage walks and community projects have all made room for the past in public life. What has changed in recent years is the scale and speed of that relationship, and the degree to which technology now mediates it.

For historians, AI is becoming more than a research tool. It is turning into part of the environment in which people encounter, interpret and remember the past. That asks us to think past the familiar question of whether AI will replace historians. The larger question is what happens to historical authority, memory and public engagement when algorithms increasingly mediate access to the past.

From the archive to the algorithm

The archive has never been a neutral repository of the past. Institutions, resources, political circumstances and cultural priorities have always shaped what is preserved, catalogued, described and made accessible. Digitisation has widened access to these collections enormously, and AI now promises to widen it further.

AI can transcribe manuscripts, recognise handwriting, extract information from large collections, translate texts, generate metadata and identify patterns across extensive archival corpora. Recent research points to its potential to make historically significant digital records more accessible, while also raising questions about appraisal, preservation, transparency and professional responsibility.

That is a real opportunity, but it brings a complication. What exists in an archive is not necessarily what becomes visible in the digital public sphere. A document may be digitised and still be hard to discover, and a source in a regional language may be poorly represented in an AI system. A community archive may lack the resources for digitisation on a large scale, while a historical narrative with substantial algorithmic visibility may come to seem more authoritative simply because it is more accessible.

The problem, then, is no longer only archival exclusion. It is increasingly a question of algorithmic visibility.

The democratisation of history

Public History should not belong only to professional historians and cultural institutions. Communities hold memories, photographs, songs, family archives, oral traditions, ritual practices and local knowledge, along with relationships to landscapes, that may never have entered a formal archive. Digital technologies have given them new ways to document and circulate this knowledge.

Research on community-generated digital cultural heritage shows how technological tools can help communities create, connect and manage their own archives. Work of this kind suggests that digital heritage becomes more participatory when communities take an active part in producing and curating their historical records, and are not merely represented by institutions. That matters most in societies with rich oral and vernacular traditions.

In India, historical consciousness has never lived only in written archives. It has also travelled through oral narratives, devotional traditions, folk performances, regional literature, songs, material culture, pilgrimage landscapes and community memory.

The challenge for digital Public History goes beyond digitising the archive. We have to ask whose archive we are digitising and whose memories are being preserved. Which languages are represented? Which communities have the technological capacity to make their histories visible? Questions like these make AI a matter of historical responsibility as well as technological innovation.

When information becomes historical knowledge

Yuval Noah Harari observes in Nexus that "information is not the same as truth." The idea reaches well beyond his book, and it is hardly new to historians. Source criticism has always required us to distinguish evidence from assertion, testimony from interpretation and documentation from memory.

What has changed is how quickly information can be assembled and presented as a coherent narrative. A generative AI system can produce a persuasive account of a historical event within seconds. Its prose may be fluent, its structure logical and its confidence considerable, but fluency does not establish historical credibility.

The historian asks different questions. Who produced the source, in what circumstances and for what purpose? What does it reveal, and what does it conceal? What other evidence confirms or contradicts it, and what assumptions shape its interpretation? These questions stay central even when the first encounter with the material comes through an AI system, and the growing use of AI makes them matter more, not less.

The problem of instant history

Part of the appeal of generative AI is that it can give an immediate answer to almost any question. History is rarely immediate. Historical knowledge is built from fragments such as documents, inscriptions, artefacts, memories, photographs, testimonies and competing interpretations, and much of the historian's work lies in negotiating those fragments rather than removing their complexity.

Recent scholarship has warned about the pull of "instant history", the expectation that AI systems can provide immediate historical answers without users engaging sufficiently with the evidentiary processes through which historical knowledge is constructed.

Here Public History faces a real tension. Public engagement requires accessibility, while historical scholarship requires complexity. Making historical knowledge inaccessible in the name of expertise is not the answer, and neither is simplifying it until every difficult question has an instant answer. The task is to build forms of Public History that are accessible without becoming simplistic and participatory without abandoning evidence.

The historian is not disappearing

Recent scholarship has begun to challenge the idea that AI's growing capacity automatically means the decline of professional historians. Chris Campbell's 2025 discussion in the Transactions of the Royal Historical Society, for instance, points to a tension. Defending historical expertise should not require historians to retreat into academic elitism or to ignore the diverse forms through which history reaches the public.

A historian's authority should not depend on holding a monopoly over historical information. It should depend on having the skills to evaluate, contextualise, interpret and communicate historical evidence responsibly.

The historian of the digital age may increasingly work as a curator who identifies significant sources, a contextualiser who places information in its historical circumstances and a verifier who tests claims against evidence. The same historian has to interpret complexity and competing meanings, mediate between scholarship and communities, and understand, as a digital humanist, the technologies through which historical knowledge now circulates.

Research on collaboration between humans and machines in historical work points the same way. AI can assist with analysis on a large scale, while human interpretation remains necessary for meaningful historical understanding.

The future is unlikely to be a simple choice between historian and machine. The more useful question is how historians can use machines without surrendering historical judgment to them.

From history about communities to history with communities

The communication of history has often followed a linear model, running from historian to institution to public. Digital Public History can make that relationship more participatory, with exchange in both directions among historian, community, archive, technology and public. Communities need not remain objects of historical representation. They can take part in documentation, interpretation and preservation.

A family photograph can join a local digital archive, an oral history can be made searchable, and a folk song can be preserved alongside its linguistic and cultural context. A community's relationship with a landscape can become part of a heritage record. AI can assist with some of these processes, but technology should not determine the meaning of the knowledge it helps preserve.

The distinction is sharpest for Indigenous and local knowledge systems. Digital preservation offers new ways to safeguard vulnerable forms of knowledge, but it can also detach knowledge from the communities, contexts and practices that give it meaning. Preservation should mean more than capturing data. It should also mean keeping context, provenance, ownership and community agency intact.

The Indian question

For India, the conversation has a further dimension. The country has great linguistic, cultural and regional diversity, and historical knowledge sits in Sanskrit and Persian manuscripts, literature in regional languages, oral traditions, folk songs, temple records, family collections, inscriptions, archaeological remains and living cultural practices. Digitisation, however, is uneven. The historical record available to an AI system is not necessarily representative of the record that exists.

If computational systems learn mainly from what has already been digitised, historical visibility can reproduce existing archival inequalities. The result may be a paradox. The technology capable of democratising history may repeat the exclusions of the archive unless the archive itself becomes more diverse.

Recent scholarship on AI and historical research in postcolonial contexts similarly identifies the marginalisation of Indigenous languages and the digital divide as obstacles, alongside the possibilities of preservation and democratisation. For Indian Public History, the future of AI cannot be separated from multilingualism, community archives, local histories and knowledge diversity.

Toward algorithmic historical literacy

This shift also changes what the history classroom is responsible for. Students increasingly need to learn to read an AI-generated historical narrative critically as well as a primary source.

They should ask where a claim came from, whether the source can be verified, what evidence supports it and what evidence contradicts it. They should ask which voices are missing, whether a complex historical process has been reduced to a simple narrative, whether an interpretation is being presented as established fact, and what assumptions may have entered the answer through the underlying data.

This could be called algorithmic historical literacy. It extends conventional historical literacy and does not replace it. Source criticism now has to work at two levels, because students must interrogate the historical source and, increasingly, the technological system through which that source reaches them.

The new responsibility of the historian

AI gives historians an occasion to reconsider what Public History is ultimately for. Taking academic history to a larger audience is only part of it. The rest is enabling people to develop meaningful, critical and informed relationships with the past.

That means leaving the confines of the archive while keeping its discipline, engaging with communities while holding to scholarly standards, and using technology without assuming that technical sophistication amounts to historical understanding. It also means recognising that historical knowledge is more than information. It is memory, identity, context, evidence and interpretation.

The historian is no longer necessarily the only person who can tell the story. The historian remains responsible, though, for asking difficult questions about how that story was constructed.

Conclusion

The most important question artificial intelligence raises may not be whether machines can write history, since they can. It is who will determine which histories become visible, which memories are preserved, which sources are trusted and which narratives enter public consciousness.

The answer cannot be left entirely to archives, institutions, platforms or algorithms. Nor should historians respond by trying to reclaim a monopoly over historical knowledge. The more productive path lies between these positions, and on it the historian of the digital age can act as a critical mediator between evidence and memory, technology and community, and scholarship and the public.

AI may help us find more sources, process larger collections and reach wider audiences. It cannot remove the ethical responsibility of deciding what should be preserved, how it should be contextualised and whose voice should be heard.

Public History should ask less whether AI can replace the historian and more whether historians can help society use AI without losing its capacity to question the past. An algorithm may generate an answer. The historian's responsibility is to teach us to ask where that story came from, whose memory it represents, what evidence sustains it and what has been left out.

Perhaps that is the historian's purpose in the age of AI, less to tell the past than to help the public learn to question the stories through which the past reaches them.

References

Campbell, C. (2025). The historian in the age of AI. Transactions of the Royal Historical Society. https://doi.org/10.1017/S0080440125100509

Hannaford, E. D., Schlegel, V., Lewis, R., Ramsden, S., Bunn, J., Moore, J., Alexander, M., Barker, H., Batista-Navarro, R., Hughes, L., & Nenadic, G. (2024). Our heritage, our stories: Developing AI tools to link and support community-generated digital cultural heritage. Journal of Documentation, 80(5), 1133--1147. https://doi.org/10.1108/JD-03-2024-0057

Harari, Y. N. (2018). 21 Lessons for the 21st Century. Spiegel & Grau.

Harari, Y. N. (2024). Nexus: A Brief History of Information Networks from the Stone Age to AI. Harper.

Jaillant, L., & Zhao, L. (2025). Introduction: When data turns into archives: Making digital records more accessible with AI. AI & Society, 40, 5787--5791. https://doi.org/10.1007/s00146-025-02374-y

Jaillant, L., & Caputo, A. (2022). Applying AI to digital archives: Trust, collaboration and shared professional ethics. Digital Scholarship in the Humanities. https://doi.org/10.1093/llc/fqac073

Teel, Z. A. (2024). Artificial intelligence's role in digitally preserving historic archives. Preservation, Digital Technology & Culture, 53(1), 29--33. https://doi.org/10.1515/pdtc-2023-0050

Yao, N. (2025). Integrating AI in historical research: A new method to analyzing national narratives. Digital Scholarship in the Humanities, 40(3), 1033--1040. https://doi.org/10.1093/llc/fqaf075

Burzlaff, J. (2025). Fragments, not prompts: Five principles for writing history in the age of AI. Rethinking History. https://doi.org/10.1080/13642529.2025.2546174

Gangwar, N. (2025). Historians: From manuscript to machine learning. Frontiers in Education, 10, 1647282. https://doi.org/10.3389/feduc.2025.1647282