Rapid developments in Artificial Intelligence (AI) are bringing increasingly complex autonomy capabilities to the cockpit. Autonomous electric Vertical Take Off and Landing, swarms, Collaborative Combat Aircraft, and other new aviation mission constructs are on the horizon. In the last few decades, military and civil aviation have achieved remarkable safety and effectiveness thanks to automation and a deliberate focus on teamwork. As automation gets replaced by autonomy, the challenges of automation could be exacerbated. Effective Human-AI teaming requires both collaborative task work and
teamwork which will be critical for continued safety and mission effectiveness. Despite the incredible ability of expert operators to make exceptional judgment calls in highly stressful situations, humans suffer from cognitive biases that may pose a challenge to this teaming. AI brings incredible data processing capabilities to the team but can suffer from a lack of adaptability to its environment and teammates, particularly in collaborative settings. As pilots retrain Crew Resource Management for their new AI mates, AI will also need to learn to adapt to its human mates. System developers can help
achieve effective human-AI teaming by providing bidirectional transparency through interface design and system features such as status, feedback, planning mechanisms, and engagement prompts.