Artificial Intelligence is becoming the key enabler of solutions to a variety of problems including those associated with cyberspace operations. Based on our analysis of cyber threats and opportunities in the coming years, we assess it as very likely that teams consisting of humans and synthetic agents will routinely work together in many if not most organizations. To fully leverage the potential of these teams, we must continue to develop new paradigms in human-machine teaming. Specifically, we must address three areas that are currently in their infancy. Firstly, we need interfaces that allow all teammates to communicate effectively with each other and seamlessly transfer tasks among them. This must be true regardless of whether the endpoints are human or not. Secondly, we will need cybersecurity operators with broad knowledge and skills. They must know how their synthetic teammates "think," when to task them and when to question their reports. Thirdly, our AI systems must be able to explain their decision-making processes to their human teammates. This paper provides an overview of cyberspace threats and opportunities in the next ten years and how these will impact human-machine teaming. We then apply the key lessons we have learned while working a multitude of advanced research projects at the intersection of human and AI agents to cyberspace operations. Finally, we propose areas of research that will allow humans and machines to better collaborate in the future.
展开▼
