Agent-based decision system for proactive cybersecurity strategies

Due to complexity and relevance for information societies, cybersecurity requires a systematic approach to threat mitigation and incident response. It must consider multi-criteria and multi-stakeholder decision-making to implement cost-effective measures. Therefore, this paper introduces a group decision system based on the Delphi technique, multi-criteria decision-making methodology, cyber mitigation framework, and integration with cybersecurity knowledge bases. Several issues are exposed related to the collaborative process, resources, and human factors, such as cognitive load and difficulty in managing complex information on cyber-attacks and protection measures. The presented research addresses these challenges by designing an autonomous agent-based decision system for proactive and reactive cybersecurity strategies. In this setting, agents utilise predictive models and represent decision-makers from multiple organisational levels and infrastructures. The paper proposes an automated collaborative consensus-seeking protocol as the central element of the agent-based system. It incorporates multi-lateral integrative negotiations, aggregation-disaggregation analysis, dynamic constraints building, coalition building, and proximity measures.