The LLM Mirage: Economic Interests and the Subversion of Weaponization Controls

Abstract

U.S. AI security policy is increasingly shaped by an $\text{LLM Mirage}$, the belief that national security risks scale in proportion to the compute used to train frontier language models. That premise fails in two ways. It miscalibrates strategy because adversaries can obtain weaponizable capabilities with task-specific systems that use specialized data, algorithmic efficiency, and widely available hardware, while compute controls harden only a high-end perimeter. It also destabilizes regulation because, absent a settled definition of "AI weaponization," compute thresholds are easily renegotiated as domestic priorities shift, turning security policy into a proxy contest over industrial competitiveness. We analyze how the LLM Mirage took hold, propose an intent-and-capability definition of AI weaponization grounded in effects and international humanitarian law, and outline measurement infrastructure based on live benchmarks across the full AI Triad (data, algorithms, compute) for weaponization-relevant capabilities.

Accepted to the ACM Conference on Fairness, Accountability, and Transparency 2026 in Montreal, Canada

References in corpus (16)

• Scaling Laws for Neural Language Models
• Emergent Abilities of Large Language Models
• Training Compute-Optimal Large Language Models
• Deep Learning Scaling is Predictable, Empirically
• Symmetry reduction for testing $k$-block-positivity via extendibility
• Outsider Oversight: Designing a Third Party Audit Ecosystem for AI Governance
• Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencent
• On the Limitations of Compute Thresholds as a Governance Strategy
• Towards an Atomic Agency for Quantum-AI
• Benchmark Early and Red Team Often: A Framework for Assessing and Managing Dual-Use Hazards of AI Foundation Models
• Data-Centric AI Governance: Addressing the Limitations of Model-Focused Policies
• Whack-a-Chip: The Futility of Hardware-Centric Export Controls
• Red Lines and Grey Zones in the Fog of War: Benchmarking Legal Risk, Moral Harm, and Regional Bias in Large Language Model Military Decision-Making
• Open-Source Assessments of AI Capabilities: The Proliferation of AI Analysis Tools, Replicating Competitor Models, and the Zhousidun Dataset
• Out of the Loop Again: How Dangerous is Weaponizing Automated Nuclear Systems?
• Quantifying CBRN Risk in Frontier Models