When an AI assistant follows the wrong instructions.
Prompt injection becomes more serious when an assistant can act. Look at the boundary between content and authority.
21 Oct 2026Read When an AI assistant follows the wrong instructions.
CYBERLENS / LOOK BENEATH THE SURFACE
THE JOURNAL / BENEATH THE SURFACE
The story behind
the screen.
The permissions you accept. The links you trust. The signals you miss. Look closer at everyday technology—and learn what to do next.
INVESTIGATIONSEXPLAINERSDEFENSIVE THINKING
PRACTICAL SECURITY / START HERE
Five lessons.
Build your judgement.
Small experiments. Readable code. A result you can explain.
Start with your own devices and follow the evidence.
01
Mobile security / 6 MIN
CODE + EXPLANATION
02Ethical testing starts with one question: what am I allowed to test?
Network security / 6 MIN
CODE + EXPLANATION
03Network threat detection: turn three failed logins into a question.
Privacy & awareness / 6 MIN
CODE + EXPLANATION
04Digital forensics: prove that your evidence copy changed.
Scripts & applications / 6 MIN
CODE + EXPLANATION
05Windows and Linux: check the firewall before changing anything.
Scripts & applications / 6 MIN
CODE + EXPLANATION
Build your first security project: a local-only header inspector.
FIELD NOTES / INVESTIGATIONS / EXPLAINERS
A clearer perspective.
IN FOCUS / AI & emerging techClear filters ×
CURIOSITY IS A DEFENCE
Don’t just read the warning.
Understand the mechanism.
Continue from the journal to a practical exercise on your own devices.