> Prompt injection is social engineering targeting reasoning entities.
That's a relief. For a moment I thought this would be about "AI".
/i
In the same way we’re making a category error in defining prompt injection, the framing of “AI agents” as primarily “intelligent actors” misses the fact that many of them will be endowed with some form of memory, be it specific to that entity or shared, and they should no longer be thought about as simply ephemeral tools.