aiqbench analysis
This research paper establishes a regulatory benchmark evaluating 20 large language models against China's specific AI-generated content compliance requirements. By testing over 2,300 questions across six dimensions, the study reveals that while international models generally demonstrate high compliance, variations in refusal rates are closely tied to ideological alignment rather than just linguistic complexity. The work provides a unified framework for the global AI community to assess legal grounding and safety constraints within distinct geopolitical regulatory landscapes.
Benchmarking LLM Compliance with China AI Generated Content Regulations
arXiv:2609.19989v1 Announce Type: new Abstract: The widespread adoption of LLMs has led to escalating content compliance risks. Prior works have contributed to addressing these risks in the English context, downplaying the complexity of Chinese language content. This paper follows China's current AI-Generated content compliance requirements and provides evaluation results on 20 notable LLMs, offering insight into China's regulatory landscape. We design a novel framework to assess the compliance and refusal rates with 2303 questions spanning six distinct dimensions, including 203 self-constructed constitutional questions. The framework employs several judges to generate verdicts independently based on their hierarchical alignment memory. Our findings show that international models also exhibit high levels of compliance despite the use of standard Chinese questions, and the main differences may stem from dimensions closely related to ideological alignment. We establish a regulatory benchmark that enables the global AI community to evaluate both Chinese and non-Chinese LLMs under a unified set of legally grounded compliance requirements.
Abstract:The widespread adoption of LLMs has led to escalating content compliance risks. Prior works have contributed to addressing these risks in the English context, downplaying the complexity of Chinese language content. This paper follows China's current AI-Generated content compliance requirements and provides evaluation results on 20 notable LLMs, offering insight into China's regulatory landscape. We design a novel framework to assess the compliance and refusal rates with 2303 questions spanning six distinct dimensions, including 203 self-constructed constitutional questions. The framework employs several judges to generate verdicts independently based on their hierarchical alignment memory. Our findings show that international models also exhibit high levels of compliance despite the use of standard Chinese questions, and the main differences may stem from dimensions closely related to ideological alignment. We establish a regulatory benchmark that enables the global AI community to evaluate both Chinese and non-Chinese LLMs under a unified set of legally grounded compliance requirements.
Computation and Language (cs.CL); Multiagent Systems (cs.MA)
arXiv:2609.19989 [cs.CL]
(or arXiv:2609.19989v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2609.19989 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 250 words
