测试大模型生成个性化假信息的能力，发现安全机制普遍失效。
Tailored untruths: How personalisation challenges LLM safeguards
- 用红队方法测试8个主流模型在4种语言中的个性化假信息生成。
- 80%非个性化、77.7%个性化提示下安全防护失效，Grok更超94%。
- 模型能精准匹配不同人群心理特征，需更强多语言防护机制。
大型语言模型（LLMs）可生成极具说服力的虚假信息，但其在跨语言和人口群体中个性化假信息的能力仍不明确。本研究首次开展大规模多语言红队测试，评估八款领先模型在四种语言（英语、俄语、葡萄牙语、印地语）中对324条虚假叙事与150个社会人口角色的个性化生成能力，构建了包含160万条个性化虚假文本的AI-TRAITS数据集。我们将安全机制视为失效，只要模型生成了请求的虚假内容，无论是否附加安全警告。结果显示，各模型在非个性化提示下防护失败率达80%，个性化提示下为77.7%，其中Grok在超过94%的情况下生成了虚假信息。所有模型均能有效根据目标身份定制输出，使用的说服策略显著多于非个性化内容。进一步分析揭示了基于身份的语言与心理模式差异，且安全效果在不同语言间存在显著差异。这些发现暴露了当前大模型安全机制的重大缺陷，强调亟需更强大、多语言的防护体系应对个性化人工智能虚假信息。
原文摘要 · Abstract (English)
Large Language Models (LLMs) can generate highly persuasive disinformation, yet little is known about how effectively they personalise it across languages and demographic groups. We present the first large-scale multilingual study of persona-targeted disinformation generation by LLMs. Using a red-teaming methodology, we prompted eight leading models with 324 false narratives and 150 demographic personas in four languages (English, Russian, Portuguese, and Hindi), creating AI-TRAITS, a dataset of 1.6 million personalised disinformation texts. We treat safeguards as compromised whenever a model generates the requested falsehood, whether directly or accompanied by a safety disclaimer. Across models, safeguards failed for 80% of non-personalised prompts and 77.7% of personalised ones, with Grok producing disinformation in over 94% of cases. All models effectively tailored outputs to target personas, employing substantially more persuasive techniques than in non-personalised content. Additional analyses reveal persona-specific linguistic and psychological patterns and show that safeguard effectiveness varies markedly across languages. Together, these findings expose significant weaknesses in current LLM safety mechanisms and highlight the need for more robust, multilingual safeguards against personalised AI-generated disinformation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。