This comprehensive survey examines the fundamental challenges that undermine AI trustworthiness, with practical insights for developing more reliable systems.

• Safety vulnerabilities including alignment issues in LLMs and harmful content generation
• Privacy concerns explored through threats like membership inference attacks
• Bias detection and mitigation strategies to ensure fair and equitable AI systems
• Security implications for developing robust AI that resists adversarial attacks

For security professionals, this research provides a structured framework to identify, assess, and address critical vulnerabilities across the AI development lifecycle.

Trustworthy AI on Safety, Bias, and Privacy: A Survey