Access-control policies, often the mechanism of choice to implement the security requirements of confidentiality and integrity, can be found in a wide range of application scenarios. Although there are standard languages for accesscontrol and a plethora of works devoted to assure the wellformedness of access-control policies, little attention has been paid to the problem of providing robust and adaptable runtime evaluation engines for the integration of accesscontrol in new DSL's and platforms. Indeed, the integration of access-control requires the development of critical infrastructure facilities around it, so that the policies can be: 1) analyzed and validated and 2) efficiently evaluated against run-time access requests.In order to solve this problem, this paper explores the use of the already mature model transformation frameworks as modern, application-independent infrastructures for accesscontrol languages i.e., following the Policy Enforcement Point(PEP)-Policy Decision Point(PDP) architecture. More specifically, we show how model-driven engineering and the ATL model-transformation framework can be used to lift the infrastructure development burden from developers by providing a robust, flexible and re-usable runtime evaluation engine for rule-based access-control policies.