Ask HN: Is your AI LLM based web service secure from prompt injection?

1 comments

• verdverm an hour ago

  I don't see how one makes an LLM safe from this, just doesn't seem like something in the realm of possibilities (without significantly nerfing the things we like about them)