Exercise 2: Why a Secret Algorithm Is a Red Flag — Possible Solution
====================================================================

Kerckhoffs's Principle states that a cryptographic system should stay
secure even if everything about it EXCEPT THE KEY is public knowledge.
A company relying on secrecy of its own algorithm's design is doing
the opposite — betting the system's security on something (the
algorithm) that the principle says should be assumed to leak sooner
or later, rather than on something (the key) that's actually designed
to be kept secret and rotated.

Why this is a red flag rather than good practice:

1. NO OUTSIDE REVIEW. A secret, unpublished algorithm has only been
   scrutinized by the small team that built it. Published algorithms
   like AES have been attacked for decades by the world's best
   cryptanalysts and have held up — that sustained public attack IS
   the evidence of real strength. A secret algorithm has no equivalent
   evidence behind it; it's untested by anyone with an incentive to
   break it.

2. SECRECY DOESN'T SURVIVE CONTACT WITH REALITY. Algorithms get
   reverse-engineered from compiled binaries, leaked by employees, or
   exposed through a data breach. The moment that happens, a system
   built on "the algorithm's design is secret" loses its only line of
   defense at once — there's no key to simply rotate, because the
   flaw was in the design itself, not a compromised key.

3. IT'S USUALLY HIDING WEAKNESS, NOT ADDING STRENGTH. In cryptographic
   engineering, refusing to publish an algorithm is treated as a
   strong signal that it wouldn't survive public scrutiny — a
   genuinely well-designed algorithm has nothing to lose by being
   published, since Kerckhoffs's Principle says its security should
   never have depended on secrecy in the first place.

The right approach — what AES, RSA, and every algorithm covered later
in this course actually do — is to publish the algorithm openly and
rely entirely on the secrecy of the key, exactly as Kerckhoffs
described in 1883.