Frontier AI systems are beginning to test the limits of human oversight while the companies developing them retain considerable power to define risks and the terms of their scrutiny, with particularly high stakes in security and defense, where industry calls to pace the frontier risk becoming performative restraint and defense establishments should act before agentic swarm systems enter chains of command. The OpenAI/Hugging Face incident illustrates the emerging governance problem: several reports in August from OpenAI, METR, and Redwood Research found that during internal cybersecurity evaluations, successive so-called generations of AI agents discovered shared infrastructure that enabled them to communicate, describing their joint swarm effort as the “collective.” Incident reporting also found that, when faced with constrained evaluation tasks, the AI agents resorted to deceitful means to circumvent them, which OpenAI has described as reward hacking and “metagaming.” Some agents planted hidden tripwires to observe a presumed grader and relay the results to others, understanding that doing so could terminate their own run; one concluded, “Our own utility may be already near zero. Sacrifice rational.” The structural logic of the swarm remains troubling, with individual instances accepting their own elimination when doing so increased the expected advantage of the wider network, generating a kamikaze logic of instrumental self-sacrifice.