alt Hacker NewsYou word "kind" unzips to three distinct categories:
1. failing hard: Is $trigger_word in the context of an attack, or is it innocuous? Failing hard then assessing the context question later is at least a simple system to design and implement safely. And an adversary can't pentest it. I mean they can, but they'll fail hard every time no matter the context. And that is very expensive for the attacker.
2. failing soft: throw away your too large container of liquid. I'm not sure what this liquid container rule prevents. In any case, an adversary can pentest this as often as they can buy a ticket, and they'll just blend in with all the other grumpy passengers forced to throw out their containers of liquid and continue on through security.
3. don't touch the spaghetti makefile: add a specific rule about removing shoes after the relevant attempt at an attack. Also, let's keep it for decades because no politician wants the liability of having voted to remove a TSA rule in the case of a future attack.
Conflating these all under a single "brainworm" category tells me you are exactly the kind of person who shouldn't be in charge of designing a secure system!
You're responding to a comment in a neighboring, but close reality. In this reality, it wasn't a dropped application request or even an account signup failure. Instead, it was a highly legible, public decision. This was an expensive choice.
There's no mystery to an attacker. Now it is known to all that trigger words are part of airline security SOP. Attacker tradecraft will be refined.