Exercise 3: Classifying Scenarios by Cryptographic Goal — Possible Solution
====================================================================

(a) An attacker reads a private email in transit.
    -> CONFIDENTIALITY. The core question is "can anyone but the
       intended recipient read this?" — and here, someone unintended
       did. No alteration or impersonation is involved, just
       unauthorized reading.

(b) A signed software update is altered after signing without
    detection.
    -> INTEGRITY. The question is "was this altered in transit?" —
       yes, and undetected alteration is exactly what integrity
       protection (hash functions, MACs — Ch.7-8) exists to catch.
       Note this scenario describes the protection FAILING, not
       succeeding — a well-implemented signature should have made
       this alteration detectable.

(c) An attacker forges a message that appears to come from your bank.
    -> AUTHENTICITY. The question is "did this really come from who
       it claims to be from?" — the message's origin is being
       impersonated, which is a different problem from whether its
       contents were readable or altered.

(d) A sender later claims they never sent a message they actually did
    send.
    -> NON-REPUDIATION. This is the one goal specifically about
       preventing a legitimate sender from denying their own past
       action — distinct from authenticity, which is about whether a
       RECIPIENT can trust who sent something, not about whether the
       SENDER can later disown it.

WHY THIS WORKS AS AN ANSWER
------------------------------
Each scenario maps to exactly one of the chapter's four goals despite
sounding similar on the surface — (b) and (c) both involve "something
is wrong with a message," but (b) is about detecting a CHANGE to
content while (c) is about detecting a FALSE ORIGIN, which is why
integrity and authenticity are treated as genuinely separate goals
requiring separate tools rather than one general "message security"
concept.