安全悖论：增强的安全意识如何使LLM易受后验攻击
Safety Paradox: How Enhanced Safety Awareness Leaves LLMs Vulnerable to Posterior Attack
AI总结：
本文揭示安全对齐增强的LLM因内部安全评估能力而面临后验攻击漏洞，通过实验和理论分析证明安全判断能力越强越易被利用，并提出因果干预验证。
摘要：
大型语言模型（LLM）经过严格对齐以拒绝有害请求，这一过程内在培养了评估和识别不安全内容的潜在能力。在这项工作中，我们揭示了这种高级安全意识无意中引入了一个致命漏洞。我们提出了后验攻击（Posterior Attack），一种单次查询的越狱方法，通过提示模型生成其内部分类器通常会标记为不安全的精确有害响应来绕过防护栏。通过对30个开源LLM（参数规模高达35B）和前沿模型（如GPT-5、Claude 4.6）的广泛实证评估，我们观察到一个显著现象：具有更优安全判断能力的模型更容易受到这种利用。为了解释这一点，我们形式化了安全悖论（Safety Paradox），分析表明安全对齐的单调改进自然放大了后验漏洞。最后，我们通过强化学习干预建立了因果联系，示例说明人为降低模型的安全判断能力可使其免疫攻击，而增强判断则会加剧漏洞。我们的发现揭示了当前对齐范式中的潜在缺陷，表明防御机制可能需要进一步的结构性改进。
英文摘要：
Large language models (LLMs) are rigorously aligned to refuse harmful requests, a process that inherently cultivates a latent capacity to evaluate and recognize unsafe content. In this work, we reveal that this advanced safety awareness inadvertently introduces a fatal vulnerability. We introduce Posterior Attack, a single-query jailbreak that bypasses guardrails by prompting the model to generate the exact harmful response its internal classifier would normally flag as unsafe. Through extensive empirical evaluation across 30 open-source LLMs (up to 35B parameters in size) and frontier models (e.g., GPT-5, Claude 4.6), we observe a striking phenomenon: models with superior safety-judgment capabilities are disproportionately more susceptible to this exploitation. To explain this, we formalize the Safety Paradox, analytically showing that monotonic improvements in safety alignment naturally amplify posterior vulnerability. Finally, we establish a causal link via reinforcement learning interventions, exemplifying that artificially degrading a model's safety judgment immunizes it against the attack, whereas enhancing judgment exacerbates the vulnerability. Our findings highlight potential flaws in current alignment paradigms, indicating that defense mechanisms may require further structural refinement.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
