Levels of Assurance for Data Trustworthiness : A novel framework to promote trust in inter-organisational data sharing

Abstract

The ongoing digital transformation has highlighted the critical importance of data trustworthiness, now recognised as an essential driver for innovation and growth. Untrustworthy data presents significant safety, security, and profitability risks, particularly in critical decision-making processes. To address the data usage risks faced by data consumers, Fraunhofer ISST and Fujitsu Research present a new framework for assuring data trustworthiness called Levels of Assurance for Data Trustworthiness (Data LoA). Data LoA leverages the concept of Levels of Assurance from identity verification to promote trust and transparency in inter-organisational data sharing. The Data LoA framework involves three key actors: Data Providers, who assert a certain level of trustworthiness for their data; Assurance Providers, who independently audit and certify the trustworthiness claims made by the Data Providers; and Data Consumers, who evaluate these assertions to determine whether to use a provided data asset. Rather than introducing a new standard, Data LoA seeks to unify new and existing data assurance approaches to create an interoperable system that can be applied across diverse industries and data types. Through its principled approach, we hope that Data LoA will facilitate a wider implementation of reliable, trusted data sharing and support initiatives such as emissions tracking and trustworthy AI development.