Yuhui Xia is studying for a Master’s at the Centre for the History of Science, Technology and Medicine at the University of Manchester and joined Archives of IT (AIT) on a placement earlier this year.
The below article on how AI can enhance the accessibility of archives is a result of that placement. Xia has also produced a high-quality report (for AIT’s internal use) that proposes AI-based solutions to improve the effectiveness and accessibility of AIT’s digital archive.
AIT has been running student placement projects for several years with the aim of showcasing how valuable our archive of oral history interviews and features is to the study of IT and digital technology.
We have done this in two ways: firstly, by asking some students to carry out research and write publishable articles on a particular aspect of the archive, and then secondly by asking others to do analytical work with our oral history interviews to support future research projects. The students’ objectives are to have some exposure to the workplace and have a discrete project that can be shown on their CVs.
Yuhui Xia has accomplished both these tasks with aplomb and AIT is extremely grateful for the commitment and professionalism shown throughout this placement.
AI could change the way we interact with archives
When I first began my research on Archives of IT (AIT) and how they use artificial intelligence, I thought AI was mainly a tool for improving efficiency such as faster searches or automatic tagging. But as I looked deeper, I realised AI might also change the way we interact with archives, making the experience more personal and engaging.
Many archives today already use AI to make information easier to find. For example, The National Archives in the UK uses AI to sort thousands of government files. Meemoo in Belgium uses facial and speech recognition to label huge amounts of video and audio. These tools help users quickly reach the materials they need. For researchers and archive staff, this means saving hours, maybe even days.
Can AI make archives more personal?
But as I explored more, I began to wonder: can AI also make the archive feel more personal, even fun?
I had this thought while browsing a website that, based on my search history, generated a kind of MBTI-style personality report. It told me things like, “Your digital personality suggests you are a curious analyst,” based on the topics I had searched for. I wasn’t expecting it—but it made me smile. It was light-hearted, but surprisingly accurate. And it made me want to explore more. That made me imagine: what if archives, especially digital ones like AIT, could offer something similar? Of course, archives are for serious research. But a small feature like this—maybe a “What kind of researcher are you?” summary—could make people feel more connected. For students or young visitors, this could be a way to feel more engaged. For researchers, maybe it helps reflect on their focus or even discover unexpected patterns in their work.
AI has power to do more than organise files
I’m not saying archives should become social media platforms. But AI has the power to do more than organise files—it can recognize patterns, respond in natural language, and recommend content. With the right balance of technology and human care, archives can become not just efficient, but also inviting.
Of course, human experts will always be important. AI systems can carry human bias from training data, and not all archives are simple. That’s why many projects—like Heritage Connector from the Science Museum Group—combine AI tools with curator input to make sure results are accurate and meaningful. In this way, AI and humans are not in competition—they support each other.
Archives work differently from country to country
As a student from China studying in the UK, I’ve become more aware of how differently archives work in different countries. In many Western digital archives, AI is used to improve user experience in creative ways. In China, archives are also changing fast, though more often focused on government use. Maybe in the future, more AI tools can be used to connect people and history in a more personal way.
In the end, archives are about memory, meaning, and people. AI can help us access the past more easily (for example, by making materials easier to find, summarizing long texts, or even offering personalised recommendations) and at the same time, it can make the experience feel more human and engaging. If AIT explored ideas like creating a digital personality for each visitor based on their interactions, it could make the archive feel even more welcoming, personal, and alive. That’s truly something worth exploring.