ClarityEthic introduces a novel framework that enables LLMs to make explainable moral judgments by emulating how humans use social norms for ethical reasoning.

• Addresses the critical safety challenge of making AI ethical decision-making more transparent and trustworthy
• Utilizes contrastive ethical insights to help LLMs understand different moral perspectives
• Enables explainable judgment rather than relying on opaque training data for moral decisions
• Aligns with human value systems by incorporating social norms into the decision process

This research significantly enhances AI security by providing a framework for ethical transparency, reducing the risk of harmful AI behaviors through explainable moral reasoning processes.

ClarityEthic: Explainable Moral Judgment Utilizing Contrastive Ethical Insights from Large Language Models