You don't necessarily need a reviewer that's immune to prompt injection. Maybe one that can express a panic state with conflicting/ambiguous material rather than going along with it could also work, and you can treat that with a shutoff to be safe, or an operator review.

Such a model doesn't yet exist though, of course.

▲

saagarjha an hour ago | parent [-]

No, you really do. Otherwise you can be prompt injected into complacency.

	▲	LoganDark 43 minutes ago \| parent [-]
		That wouldn't really fit what I just described at all. Obviously with current architectures, higher resistance to prompt injection is the best you can do.