Recommending an encryption software by Paranoia Works

JohnDose:

It’s just an option afterall

It’s a footgun. Cryptography lives from stuff like best practices, opinionated crypto, and also, misuse resistance. Since it’s not safe to use and requires blind trust on every session, offering it in the first place shows incompetence and when paired with usual novice mistakes it doesn’t inspire confidence in the project overall.

Cryptography is the structural integrity engineering of digital infrastructure. You don’t want amateurs designing or building bridges, or designing materials used to build them. So you don’t want amateurs designing the ciphers, implementing them, or writing protocols and software that uses them.

undiscovered attacks using quantum computers

It’s much more likely there’s an undiscovered attack against RC6 or Blowfish because these ciphers aren’t being analyzed anywhere nearly as much. For quantum computers, the best attack we know is Grover’s algorithm, and it can’t break even AES-128 in practice.

it isn’t flawless

Sounds like you’re trying to market the cascading encryption indirectly. Again, if ChaCha20 or AES is broken and the generated random key is predicatable, it doesn’t matter which algorithm you use. And no matter what the system will rely on ML-KEM for key exchange, that’s another single point of failure. It’s much more probable these new post-quantum kex algos have problems than extensively analyzed symmetric ciphers. I’ve never seen concerns towards AES or ChaCha.

I know nothing about cryptography

I think this says more than the “I also believe it isn’t flawless”. AES isn’t flawless, but the interesting part is how big of a dent has the entire academic community managed to make. The Biclique attack is the best classical one yet, and it has managed to reduce key size by a whopping 1.6 bits. So AES256 is now AES254.4. The standard rule of cryptography is that attacks only get better, but quarter of a century later, Biclique attack is like a paint scratch in an aircraft carrier’s hull.

That’s why I was looking for alternative encryption algorithms, preferably with larger key sizes.

Again, nobody is breaking AES ciphertexts with some secret quantum algorithm. Any attacker will have a budget and to them the bang for buck is currently in “Let’s put Claude Mythos find a zero day in the operating system, and hack the computer / phone and watch the target type the message and exfiltrate the plaintext or the key(s).” Finding the zero day costs maybe 50,000 USD, after which it can pwn targets for 1-120 months until it’s detected and patched. That’s fractions of a penny compared to the crypto nerd’s fantasy.