Adaptivity has become a prominent research topic during the past decades. There is a variety of application areas where adaptation can add benefit to systems, reaching from shopping portals to e-learning platforms. Despite the large body of work, Computational Intelligence (CI) techniques have been under-explored and therefore also under-exploited within the area of useradaptive systems. This paper discusses the potential of employing CI approaches for the implementation of adaptivity within e-learning systems. Furthermore, it identifies scenarios where these techniques can improve the performance of an adaptive component. Specific focus is placed on the provision of guidance in e-learning systems, in particular with respect to communication/cooperation (as opposed to traditional focus on guiding learners through learning materials).