The traditional learning paradigm invoving face-to-face interaction with students is shifting to highly data-intensive electronic learning with the advances in Information and Communication Technology. An important component of the e-learning process is the delivery of the learning contents to their intended audience over a network. A distributed learning system is dependent on the network for the efficient delivery of its contents to the user. However, as the demand of information provision and utilization increases on the Internet, the current information service provision and utilization methods are becoming increasingly inefficient. Although new technologies have been employed for efficient learning methodologies within the context of an e-learning environment, the overall efficiency of the learning system is dependent on the mode of distribution and utilization of its learning contents. It is therefore imperative to employ new techniques to meet the service demands of current and future e-learning systems. In this paper, an architecture based on autonomous mobile agents creating a Faded Information Field is proposed. Unlike the centralized information distribution in a conventional e-learning system, the information is decentralized in the proposed architecture resulting in increased efficiency of the overall system for distribution and utilization of system learning contents efficiently and fairly. This architecture holds the potential to address the heterogeneous user requirements as well as the changing conditions of the underlying network.