101922Digital FingerprintingClosedLaunchpadInnovate UKCurrent commercial tools aimed at data loss prevention are mostly rule based; inspecting data against a set of rules and choosing to block (alert) or allow like a firewall. The rules are content or context based and require significant effort to develop and maintain bespoke rules to protect a specific organisations assets. This research has developed a digital fingerprinting technique that uniquely identifies the content of classified documents stored within a network, based on the content itself and not the presentation or encoding of the file in which it is stored. The technique allows the digital fingeprinting of a range of file formats, including images and source code that can be used to detect attempts to move sensitive data, either as while documents or fragments of documents, out of a network.