"DARPA Accepting Proposals for CASTLE Program to Fortify Computer Networks"

When defending critical computing assets, an everchanging cyberattack surface, infrequent computer vulnerability scans, and burdensome security procedures create an uneven battle. When those factors are combined with the high cost of cybersecurity assessments that often fail to provide actionable feedback, the odds may favor malicious actors. The Defense Advanced Research Projects Agency (DARPA) plans to change that dynamic by launching a new technology-focused program that can speed up cybersecurity assessments through automated, repeatable, and measurable approaches. The Cyber Agents for Security Testing and Learning Environments (CASTLE) program aims to improve cyber testing and evaluation by creating a toolkit that makes realistic network environments and trains Artificial Intelligence (AI) agents to defend against Advanced Persistent Threats (APTs). Teams will use Reinforcement Learning (RL), a type of Machine Learning (ML), to automate the process of reducing network vulnerabilities. Attackers are often found to understand network vulnerabilities better than defenders, but this does not have to be the case, according to Tejas Patel, CASTLE program manager in DARPA's Information Innovation Office. RL could enable the development and training of cyber agents that are significantly more effective than current manual approaches to dealing with APTs in networks. This article continues to discuss the goals of DARPA's CASTLE program.  

HSToday reports "DARPA Accepting Proposals for CASTLE Program to Fortify Computer Networks"