<p>This study investigated how users perceive the concept of “knowledge” in robots and how this perception differs from their views on human knowledge. Based on the Justified True Belief (JTB) framework for knowledge, I conducted three experiments to verify whether users apply the same standards to a robot’s knowledge as they do to human knowledge. In Experiment 1, I demonstrated that the conditions under which users attribute knowledge differ between humans and robots through an experiment where participants read a Gettier-type scenario and answered questions. Experiment 2 examined the threshold at which information is perceived as “knowledge,” revealing that users tend to demand more direct verification from robots than from humans. Experiment 3 examined the impact of beliefs about the robot’s knowledge, identified in Experiments 1 and 2, on users’ moral judgments. The results showed that users attributed less responsibility to the robot as an accomplice to the crime than to a human accomplice. These results suggest that users perceive robots’ knowledge as inflexible, requiring direct perception, and fundamentally different from human knowledge and reasoning. These findings provide important insights for the design of human–robot interaction (HRI) systems. Understanding users’ perceptions of robot knowledge may be essential for building trust in future human–robot collaboration.</p>