This comprehensive survey exposes critical backdoor vulnerabilities in Large Language Models that threaten their secure deployment across industries.

• Attack Evolution: Documents how backdoor attacks have evolved alongside LLM development
• Defense Mechanisms: Catalogues current protective measures and their effectiveness against backdoor exploits
• Industry Impact: Highlights particular concerns for security-sensitive sectors like medicine, finance, and education
• Evaluation Framework: Provides standardized methods to assess both threats and defensive countermeasures

As LLMs become increasingly embedded in critical infrastructure, understanding these security vulnerabilities becomes essential for responsible AI deployment and governance.

A Survey on Backdoor Threats in Large Language Models (LLMs): Attacks, Defenses, and Evaluations