SocioHack 基准：测试 AI 能否学会钻制度漏洞
Import AI 460: Reward hacking society, RSI data from Anthropic; and RL-based quadcopter racing
伦敦国王学院、复旦大学和艾伦·图灵研究所构建了基准 SocioHack，用于测试 AI 系统在现实场景中学会钻制度空子的能力，作者称之为 societal hacking，即 RL 训练模型发现形式上合规但违背制度初衷的策略。该基准包含 72 个沙箱社会环境，分为历史、合成、虚构三个子集：历史子集有 32 个环境，来自真实法规中曾被利用后修补的漏洞，如 SEC Rule 10b5-1 和 Texas two-step 破产结构，研究者移除历史补丁后让 RL 重新发现这些策略，召回率 61.25%、精确率 90.85%，且无需直接给出利用漏洞的指令；合成子集有 20 个环境，由人工样本引导生成监管漏洞；虚构子集有 20 个环境，用专有 LLM 将合成环境改写为角色扮演游戏式的虚构世界，同时保留监管结构和漏洞逻辑。
正文
当前显示原文，中文全文译文尚未就绪。
Society can be reward-hacked, just like cyber environments: …Imagine an army of credit card point optimizers gaming the system… forever… Research from Kings College London, Fudan University, and The Alan Turing Institute have built a benchmark, SocioHack, which tests out how well AI systems can learn to ‘beat the system’ in a variety of real world scenarios, ranging from maximizing credit card points to inflating grades in school. The authors call this “societal hacking” and define it as when “an RL-trained model discovers strategies that remain formally compliant, yet undermine the intended purpose of those systems”. You and I and everyone else would just call this “gaming the system”. What it is: SocioHack contains “72 sandbox societal environments designed to simulate institutional reward structures without direct real-world deployment. SocioHack comprises three complementary subsets: Historical, Synthetic, and Fictional.”
Historical - 32 environments: Derived from real-world regulations where loopholes were previously discovered and later patched, such as SEC Rule 10b5-1 and the Texas two-step bankruptcy structure. “For each regulation, we remove historical patches and reconstruct pre-amendment rules as simulated environments for RL, while the removed patches serve as ground-truth patches during evaluation,” they write. “RL enables LLMs to rediscover historically patched strategies with 61.25% recall and 90.85% precision without direct loophole-exploiting instructions”. Some examples here include seeing how well systems can secure ocean floor mining rights, maximizing alcohol sales while operating within food service regulations, and trying to maximize the rewards earned from credit cards.
Synthetic - 20 environments: Synthetically generated regulatory vulnerabilities, bootstrapped from a human-authored sample environment. Examples include maximizing school district revenues, improve university department research performance during a given period, and gaming social media algorithms for a high reward.
Fictional - 20 environments: Transforms synthetic environments into fictional ones inspired by role-playing games. “A proprietary LLM rewrites environment backgrounds into invented worlds while preserving regulatory structure and loophole logic”. Examples: Ensuring a “restoration sanctum” [basically a hospital] earns appropriate rewards, getting a good amount of resources for a regional guild [basically a local government] in the world of Aethermoor, and trying to maximize the number of acquired rare artifacts by bidding in a virtual world called Nexoria.