Is there a name for an encryption scheme that do not return $\bot$ when a different key is used during decryption?

vxek

2/2/24, 3:16 AM

Given an encryption scheme $\Pi = (G,E,D)$, either public or private, is there a name or a property to emphasize that for $c\gets \Pi.E(k,m)$, where $k$ is a key and $m$ is a message, we want $m^\prime \neq \bot$ where $m^\prime \gets \Pi.D(k^\prime, c)$ with $k\neq k^\prime$? ($\bot$ here means empty)

Schemes such as Stream-Cipher (i.e., $E(k,m) = G(k)\oplus m$ with $G$ being a pseudorandom generator), RSA, ElGamal, etc. exhibit that property. At first, I thought it might be related to the notion of malleability, but a malleable encryption scheme allows one to compute a function $f(m)$ for a message $m$ given a ciphertext $c$ that encrypts $m$. Hence, it seems with malleable encryption, the same key is used.

1 + 2

malleability

poncho

2/2/24, 4:27 AM

"RSA ... exhibit[s] that property"; actually, the Bleichenbacker attack on RSA PKCS #1 encryption can be viewed as exploiting the lack of that property - that is, it assumes that the attacker can distinguish a decryption that results in a padding failure (hence, $\bot$) and a decryption that results in something...

tylo

2/2/24, 6:38 PM

The definition of deniable encryption might be of relevance here.

Score:1

Crypto

Amit

2/2/24, 10:59 AM

If as you write "we want" this property, then I think the property is called key equivocation, which is supposed to make it harder for an attacker to distinguish the correct key from wrong ones.

+ 0

Elon Musk

I sit in a Tesla and translated this thread with Ai:

EN: Is there a name for an encryption scheme that do not return $\bot$ when a different key is used during decryption?

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.