SAIF: A Comprehensive Framework for Evaluating the Risks of Generative AI in the Public Sector