arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2403.03550cs.AIcs.CYcs.HC
收藏

通过提示工程进行的情绪操纵会放大AI大语言模型的虚假信息生成能力

Emotional Manipulation Through Prompt Engineering Amplifies Disinformation Generation in AI Large Language Models

University of Zurich(苏黎世大学);University of St. Gallen(圣加仑大学)

Rasita Vinay, Giovanni Spitale, Nikola Biller-Andorno, Federico Germani

AI总结:

本研究测试多款OpenAI大语言模型的虚假信息生成能力,发现情绪性提示会影响其生成频率,礼貌提示下生成率高,不礼貌提示下会拒绝生成,为AI负责任应用提供参考。

摘要:

本研究探究了OpenAI的大语言模型(LLMs)通过提示工程生成合成虚假信息的情况,并考察了这类模型对情绪性提示的响应能力。我们利用davinci-002、davinci-003、gpt-3.5-turbo和gpt-4等多个LLM迭代版本设计实验,评估其生成虚假信息的成功率。基于包含19800条合成虚假信息社交媒体帖子的语料库,我们的研究结果显示,OpenAI的所有LLM都能成功生成虚假信息,且能有效响应情绪性提示,表明它们在文本生成中对情绪线索具备精细化理解能力。当使用礼貌语气提示时,所有受测LLM都持续以高频率生成虚假信息。反之,当使用不礼貌语气提示时,虚假信息的生成频率会降低,因为模型通常会拒绝生成虚假信息,反而会提醒用户该工具并非为此类用途设计。本研究为围绕AI技术负责任开发与应用的持续讨论作出了贡献,尤其是在遏制虚假信息传播、提升AI生成内容透明度方面。

英文摘要:

This study investigates the generation of synthetic disinformation by OpenAI's Large Language Models (LLMs) through prompt engineering and explores their responsiveness to emotional prompting. Leveraging various LLM iterations using davinci-002, davinci-003, gpt-3.5-turbo and gpt-4, we designed experiments to assess their success in producing disinformation. Our findings, based on a corpus of 19,800 synthetic disinformation social media posts, reveal that all LLMs by OpenAI can successfully produce disinformation, and that they effectively respond to emotional prompting, indicating their nuanced understanding of emotional cues in text generation. When prompted politely, all examined LLMs consistently generate disinformation at a high frequency. Conversely, when prompted impolitely, the frequency of disinformation production diminishes, as the models often refuse to generate disinformation and instead caution users that the tool is not intended for such purposes. This research contributes to the ongoing discourse surrounding responsible development and application of AI technologies, particularly in mitigating the spread of disinformation and promoting transparency in AI-generated content.

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。

↑