Approaches for Using AI to Manage Records at Scale

27th June 13:10 – 13:30

Speaker: Dr James Lappin

Abstract: This presentation explores pathways by which an originating organisation could apply AI to managing digital records at scale. It asks which pathways are likely to offer the most traction to the widest range of organisations over the widest range of content repositories.

The collaboration and communication tools in common use in a typical organisation in the early 2020s have a modular architecture. Such systems are typically deployed by provisioning accounts to each individual, and sites to each team. This modular architecture is not optimised for the efficient management of records over time, but is very effective in supporting the speedy deployment of new tools.

Records management is concerned with the efficient, predictable, precise and reliable application of retention and access rules to records. Rules are applied more predictably if they are applied to aggregations of items rather than to individual items. Records management theory holds that rules are applied more precisely if records are aggregated by business activity. However theory also asserts that the original order within which records were accumulated should be respected, even if it is imperfect. This is because that order constitutes an important part of the context of any given record.

AI will offer the prospect of reorganising any repository of content, at any stage of the records lifecycle. This talk explores ways of managing the tension between using AI to improve the way records are organised, and preserving the context of records, including the way they were originally aggregated.

Bio: Dr James Lappin is a Senior Policy lead for knowledge, information and records management at the Central Digital and Data Office. His PhD thesis The Science of recordkeeping systems – a realist perspective was published earlier this year. James is the author of the Thinking Records blog.