Fully automated demand response may be implemented at end users, in accordance with terms agreed to by end users to reduce energy demand during demand response events. Demand reduction actions to implement the objectives of a demand response event at the end users may be determined, desirably using artificial intelligence and neural networks, based on energy demand curtailment objectives of the demand response event, hierarchy(ies) of demand reduction actions for respective demand response events ordered to minimize undesired impact at the end users, and monitoring data received from, or relating to implementing energy demand curtailment at, the end users. In addition, demand reduction actions may be automatically implemented at end users in the absence of a demand response event, to implement energy demand curtailment according to criteria of end users, where the demand reduction actions are determined based on monitoring data and a hierarchy(ies) of demand reduction actions and using artificial intelligence and neural networks.