Recently, many teams are working on Text-to-SQL, ChatBI, or data analysis Agents. One underestimated issue is that SQL generated by LLMs should not directly enter production databases. This article discusses: explaining LLM SQL Guard with clear definitions: why Text-to-SQL cannot rely solely on model generation, and why deterministic SQL checks are mandatory before execution. Key points: LLM-generated SQL may be syntactically correct but semantically incorrect or unsafe. SQL Guard performs deterministic checks before execution by combining parser, catalog, policy, risk, and audit. Suitable as an introductory article for AI data governance, ChatBI, and Text-to-SQL teams. Original link: dpriver.com/blog/what-is-an-llm-sql-guard