Using AI to Manage the Digital Heap

The digital heap consists of the entirety of the unstructured content that organisations are holding in their live and legacy collaboration systems, document repositories and email systems. This content is likely to spread across multiple applications, each with its own distinct information architecture.

This talk covers the various stages involved in designing and carrying out an AI intervention for records management purposes. It sets out the range of different options open to an organisation as they decide what content to target, at what stage of its lifecycle, with what AI tools/data science techniques, for what objectives, and with what governance arrangements.

The talk draws on the Government Digital Service’s AI Insights guide ‘Using AI to manage the digital heap’ which was published in September 2025. The guide explains how AI (and data science techniques) can be integrated into an organisation’s overall approach to applying retention rules to content.