A Model-Based Privacy Compliance Checker

The Need for Automation

We address the problem of how to make privacy management more effective by introducing more technology and automation into the operation of privacy in e-business organizations. Enterprises are coming under increasing pressure to improve privacy management, both to satisfy customers and to comply with external regulation (Laurant, 2003) or internal policies. Not only are human processes prone to failure but the scale of the problem highlights the desire for additional technology to be part of the solution. The trend towards complexity and dynamism in system configurations heightens this need for automation to ensure that privacy and security properties are maintained as changes occur, and in addition to check that the privacy enhancing technologies are operating as desired.

Automated Compliance Checking Requirements

Most of the technical work done in this space focuses on the provision of auditing and reporting solutions that analyse logged events and check them against privacy policies and process guidelines. These auditing systems usually operate at a low level of abstraction and do not take into account the overall compliance management process that involves both the refinement of privacy laws and guidelines within enterprise contexts, their mapping into the enterprise IT infrastructure and their subsequent checking against the enterprise’s operational behaviour.

At present there is a gap between the definition of high-level regulations, standards and best practices and what is actually happening in an enterprise at the level of application software, system software and middleware, processors, networks and data stores. The current approach is to fill this gap using people-based processes, but there are drawbacks to this, in terms of being slow, expensive, error-prone and leading to best-effort compliance due to limited resources. Our vision is to bridge this gap where possible with model-based technology and automation, as shown in Figure 1. On the one hand privacy policy enforcement technologies can be used to deliver compliance to privacy principles and goals; on the other hand (the focus of this article) we can use system monitoring technologies to continuously assess their actual performance and ability to deliver against the requirements of the policy.

Figure 1.

Model-based, policy-driven IT