🤖 AI Summary
This study addresses a critical gap in AI governance by demonstrating that the prevailing focus on high-compute models overlooks emerging safety risks from low-compute models, whose capabilities have surged due to algorithmic advances. Through systematic analysis of over 5,000 open-source large language models, combined with historical benchmarking, parameter quantization, resource simulation, and deployment experiments on consumer-grade hardware, we show that model compression and agent-based workflows are enabling hazardous capabilities—such as disinformation generation and voice-cloning fraud—to migrate efficiently to lightweight models. Our findings reveal that the model scale required to match mainstream LLM performance has decreased by more than an order of magnitude within a year, rendering many digital-society harms executable on everyday devices. This challenges compute-centric regulatory paradigms and underscores the urgent need for governance frameworks that account for low-compute AI risks.
📝 Abstract
Artificial intelligence (AI) systems are revolutionizing fields such as medicine, drug discovery, and materials science; however, many technologists and policymakers are also concerned about the technology's risks. To date, most concrete policies around AI governance have focused on managing AI risk by considering the amount of compute required to operate or build a given AI system. However, low-compute AI systems are becoming increasingly more performant - and more dangerous. Driven by agentic workflows, parameter quantization, and other model compression techniques, capabilities once only achievable on frontier-level systems have diffused into low-resource models deployable on consumer devices. In this report, we profile this trend by downloading historical benchmark performance data for over 5,000 large language models (LLMs) hosted on HuggingFace, noting the model size needed to achieve competitive LLM benchmarks has decreased by more than 10X over the past year. We then simulate the computational resources needed for an actor to launch a series of digital societal harm campaigns - such as disinformation botnets, sexual extortion schemes, voice-cloning fraud, and others - using low-compute open-source models and find nearly all studied campaigns can easily be executed on consumer-grade hardware. This position paper argues that protection measures for high-compute models leave serious security holes for their low-compute counterparts, meaning it is urgent both policymakers and technologists make greater efforts to understand and address this emerging class of threats.