The automated negotiation process seems to be a powerful mechanism to resolve disputes arising from Internet-based transactions. Automated negotiation is an online blind-bidding process in which an automated algorithm evaluates bids from the parties and settles the case if the offers are within a prescribed range. Following the arguments of the dispute resolution professionals, the main advantage of this procedure is to promote “natural” agreements by restoring the parties' right to negotiate on their own, without the presence of a third party in the shadow of negotiations. Our purpose is to investigate this issue by modelling the automated negotiation process as a two-person bargaining game under incomplete information. A first result states that, given incomplete information, not all mutually beneficial agreements can be attained via the procedure. Furthermore, the settlement rule has a drastic effect on the players' strategies, which induces that the automated negotiation process does not significantly increase the likelihood of a settlement. The ability of the procedure to generate efficiency is only due to the costs imposed on parties if a disagreement occurs, that is the combination of players' risk aversion and uncertainty.