The Design Council Archive at the University of Brighton Design Archives provides insight into the history and development of design in the mid-20th century in the United Kingdom and internationally. One of the strengths of this archive is its photographic library; a collection of over 100,000 prints and 60,000 negatives used by the Design Council (previously the Council of Industrial Design) to promote ‘good’ design. It was used as a marketing tool for manufacturers, journalists and buyers, to support teaching design in the education sector, in addition to being a general design aide and resource for the general public. One subset of this collection is a series of over 11,000 black and white glass-plate negatives, digitised over five years, and are identifiable by a sequential list of image numbers. A set of 26 registers organised chronologically provide the content and structure for this data set including manufacturer, copyright, accession date, and taxonomic category used by the Council to organise its photo library. The sequential numbering does not correspond to the chronological order, making it virtually impossible to identify the content in each negative, nor where it sits in the Council’s taxonomy. This paper explores the different phases of a project that employed Artificial Intelligence (AI) and Machine Learning (ML) to create a tool to make this data set of 11,000 uncatalogued glass-plate negatives accessible and usable, including discussion of initial outcomes, challenges presented, as well how AI/ML can be further employed to diversify stakeholder representation and increase access to archives.

This paper presents various innovations at the National Archives of Singapore (NAS), to improve recordkeeping for Singapore’s Whole-of-Government and promote wider awareness of our records to the public.

The following initiatives implemented throughout the entire records lifecycle will be highlighted:

Firstly, NAS conducted a pilot Artificial Intelligence (AI) trial with Singapore’s Government Technology Agency to improve appraisal turnaround. The trial focused on mapping records series to pre-approved Records Retention Schedules for common corporate records. Details about the process of planning and ideation, learning points, and areas of limitations will be covered.

Next, new digitisation approaches for agencies’ digitalisation and eventual transfer of records to NAS. NAS has established a bulk tender with a panel of vendors, which government agencies can leverage to digitise records of various formats (excluding audiovisual) as part of the Government’s digitalisation move. NAS is also planning a Centralised Secure Digitisation Centre, to address government agencies’ need for digitisation of classified records.

Lastly, NAS creatively publicised and promoted awareness of our records by engaging the public via social media platforms e.g. Facebook, Instagram and Tik Tok, to showcase our collections and work. NAS also collaborates with the GLAM sector annually during the Singapore Night Festival (SNF) and Love Our Libraries and Archives (LOLA) event to feature selected collections and invite the public to discover the Archives in person.

Artificial Intelligence is revolutionizing the way archives and records are accessed, preserved, and utilized. This presentation explores innovative AI-driven solutions that address pressing challenges in recordkeeping, showcasing their relevance to the ARA 2025 theme, ‘Next Generation: Innovation and Imagination in Record Keeping.’

Drawing on real-world projects, including collaborations with national archives, the session will delve into two transformative applications: visual search and automated information extraction. Visual search empowers users to intuitively explore large collections of images and videos without exhaustive metadata, while automated extraction unlocks insights from unstructured records, making inaccessible information discoverable and actionable.

The presentation highlights practical approaches that reduce manual workloads, enhance accessibility, and address the sustainability and inclusivity pillars central to the sector’s future. Ethical considerations, such as bias, data transparency, and the archivist’s role in responsible AI adoption, will also be discussed.

Participants will gain a non-technical understanding of these technologies, practical examples of their implementation, and insights into how they can be adapted to diverse archival contexts. By bridging traditional practices with cutting-edge innovation, this session aims to inspire archivists to reimagine their roles and embrace the opportunities AI offers for shaping the next generation of recordkeeping.