This research explores how Large Language Models can generate accurate, secure cloud access control policies based on natural language requests.

• LLMs can effectively translate human specifications into formal access control policies
• The approach reduces manual policy writing errors that often plague security configurations
• Models can be fine-tuned to improve accuracy and security alignment
• The system provides a practical solution for cloud administrators managing complex security rules

Security Impact: By automating policy creation, organizations can strengthen their security posture while reducing the risk of human error in critical access control configurations - a leading cause of security breaches in cloud environments.

Synthesizing Access Control Policies using Large Language Models