Scholarlink

<p>Much Large Language Model (LLM) use is happening undisclosed. This is troubling because the undisclosed use of LLMs to mediate interpersonal communication can be a morally objectionable kind of deception. It is morally objectionable when the intentional failure to disclose LLM use deprives others of valuable information they need for their epistemic appraisal of, <i>inter alia</i>, another’s authentic competencies. In this paper, we provide an analysis of deception in the context of undisclosed LLM use and argue that in some cases this amounts to an ethically salient epistemic wrongdoing.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Is Undisclosed LLM Use Morally Wrong?

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月24日

查看全文
  • Siavosh Sahebi,
  • Thomas Montefiore

摘要

Much Large Language Model (LLM) use is happening undisclosed. This is troubling because the undisclosed use of LLMs to mediate interpersonal communication can be a morally objectionable kind of deception. It is morally objectionable when the intentional failure to disclose LLM use deprives others of valuable information they need for their epistemic appraisal of, inter alia, another’s authentic competencies. In this paper, we provide an analysis of deception in the context of undisclosed LLM use and argue that in some cases this amounts to an ethically salient epistemic wrongdoing.