White paper
In an internal investigation, the decisive difficulty often lies in the relationships between documents: the same word — such as ‘approval’, ‘validation’ or ‘control’ — may refer to different actions, conditions or authorities. The challenge is to use AI to make these relationships open to examination without removing context, uncertainty or professional responsibility.
Paper podcasts
Free access. No registration.
The paper explains how to compare objects, points in time, versions and conditions across multiple sources. It emphasises the distinction between statements, instructions, application records and visual indicators so that each connection remains interpretable.
Connections between sources are used to formulate checks and interview questions whose potential effect on the next steps can be identified. New evidence can therefore be linked to the relevant questions without indiscriminately reopening the entire case.
The white paper describes a working record linking findings, sources, actions and unresolved points to support review and case handover. This approach is complemented by criteria covering authorised access, document coverage, omissions and evaluation in relation to specific uses.
This white paper explains how to connect emails, invoices, interviews, procedures, application logs and other formats around the questions they help to address. It considers how to explain these connections, distinguish their supporting evidence and turn them into checks or interview preparation with a clear purpose. It also shows how to incorporate new evidence by identifying what it confirms, clarifies or changes, and how to prepare audience-appropriate reports that remain linked to their sources. Finally, it proposes criteria for use and evaluation covering scope, access rights, omissions and the conditions required for confidence.
Internal investigation leaders and teams; compliance and ethics functions; legal departments; internal audit teams; authorised advisers involved in investigations; data protection officers; and teams responsible for the governance or evaluation of AI systems.
