1 year ago
Fri Jan 3, 2025 8:51pm PST
Ask HN: How to Safeguard a LLM to SQL Prompt from SQL Injection?
I have seen posts about how you can use LLM to proxy queries into your dataset. This seems powerful, but how to prevent it from being abused or hacked? Are there other strategies to deploy a text to SQL use case?
comments:
add comment
loading comments...