The automatic prevention, detection, and reaction for intrusion management has been a key issue for years, focusing on the use of IDS-based approaches. In addition, dynamicity and the changing nature of the technology and threats have led to consider other approaches. In this paper, we present two R&D projects whose purposes are addressing the above shortcomings. First, we present the RECLAMO project, where an architecture for an Automated Intrusion Response System is proposed to divert a given attack to a honeynet, dynamically built based on the attack information. Secondly, we also describe an ongoing R&D project, called DHARMA, where an efficient Dynamic Risk Assessment and Management is proposed to measure the risk level on the organization's assets at real time, taking the required actions as a response from a proactive defense model.