Exchange of arguments in a discussion often makes individuals more radical about their initial opinion. This phenomenon is known as Group-induced Attitude Polarization. A byproduct of it are bipolarization effects, where the distance between the attitudes of two groups of individuals increases after the discussion. This paper is a first attempt to analyse the building blocks of information exchange and information update that induce polarization. I use Argumentation Frameworks as a tool for encoding the information of agents in a debate relative to a given issue a. I then adapt a specific measure of the degree of acceptability of an opinion (Matt and Toni 2008). Changes in the degree of acceptability of a, prior and posterior to information exchange, serve here as an indicator of polarization. I finally show that the way agents transmit and update information has a decisive impact on polarization and bipolarization.