Prompt injection
An attacker getting instructions into your context that your model then follows as if they came from you.
By Lattice editorial · Entry last reviewed · method
What it implies
Not a bug to be fixed but a property of systems that read untrusted text and act on it. Input filtering catches known patterns; the durable mitigation is not trusting the model's reasoning about permissions, and re-deriving authorisation from authoritative state rather than accepting its claim.