Cognitive AI systems are hindered by significant gaps in their capability to reason, adapt, and self-regulate, despite advancements in generative and agentic AI. These limitations restrict reliable operation and hinder the development of systems that can sustain complex tasks. Researchers have identified key cognitive functions that remain underdeveloped, including persistent memory and adaptive behavior. The lack of these fundamental capabilities undermines the potential of cognitive AI to move beyond language generation and autonomous task execution. This shortfall has significant implications, particularly in the context of state-aligned threat activity, where the stakes extend beyond the immediate target to the geopolitical sphere1. The development of more sophisticated cognitive AI systems is crucial to addressing these gaps and mitigating potential threats. So what matters to practitioners is that these capability gaps must be prioritized and addressed to ensure the reliable operation of cognitive AI systems in high-stakes environments.