Watch this video on YouTube
An NLP application is being secured against prompt injection. Which method would best validate incoming inputs to prevent exploitation?