Purposeful or malicious harm

Mitigation strategy

1. **Implement and Enforce Human-in-the-Loop (HITL) Protocols for Lethal Action** Prioritize establishing and mandating robust governance frameworks, such as the Human-in-the-Loop (HITL) or Human-on-the-Loop (HOTL) models, across all deployed EAI systems with potential for lethal or mass harm. This involves setting clear international and national policy limits on the types of targets and scope of engagement to ensure human operators retain ultimate, meaningful judgement and control over the use of force throughout the weapon system's lifecycle, thereby mitigating the risk of intentional or unintentional misuse by malicious actors. 2. **Enhance System Resilience and Assurance through Rigorous V\&V** Require and implement rigorous verification and validation (V\&V) protocols, along with comprehensive safety engineering design reliability, to ensure EAI systems operate as intended and minimize susceptibility to exploitation. This includes cyber-hardening the autonomous platforms via strong encryption, security measures to prevent model and data theft from edge devices, and conducting adversarial testing to proactively address vulnerabilities to hacking, spoofing, and other forms of malicious input or compromise. 3. **Develop and Integrate Advanced Counter-EAI Defense Systems** Invest in and deploy sophisticated, multi-layered Counter-Unmanned Aerial Systems (C-UAS) and counter-EAI technologies that are capable of detecting, tracking, and neutralizing malicious autonomous platforms. Mitigation strategies should emphasize non-kinetic "soft-kill" techniques, such as cyber-based takeover, signal manipulation, and targeted jamming, to safely neutralize the threat and enable forensic recovery, which is critical for non-state actor or non-permissive environment scenarios.