e address the task planning problem for robots collaborating with humans to achieve a shared task. In this context, humans often want to be free to choose \textit{what} they want to do and \textit{how}, according to their internal preferences. We propose a concurrent and compliant joint action model based on social and collaborative signals. This model captures subtle possible agents' coordination and the human's inherent uncontrollability. We use an abstracted version of this model to explore relevant courses of action and guide our planning approach. The latter produces a behavioral policy that describes the robot's action to perform concurrently in every state, such that the action is congruent and compliant with any online human's decision and action, including being passive. The policy also aims to satisfy, at best, an estimation of the human internal preferences, which can be updated online. In this paper, we present our proposed joint action model and describe a novel human-aware task planning approach based on this model. Then, using BlocksWorld scenarios, we provide empirical evidence depicting the compliant concurrent robot behaviors produced by our approach. Additionally, we implemented our joint action model as an execution scheme into an interactive simulator used to conduct a user study. Compared with a baseline where robot decisions are always imposed on humans, our approach was significantly more appreciated and better satisfied the participants' preferences.