In a standard asymmetric-key (public-key) encryption exchange -- where a…

Asymmetric-key (public-key) cryptography gives every participant a linked pair of keys: a public key, which is published for anyone to use, and a private key, which its owner alone keeps secret and never transmits to anyone else. Confidentiality for a message is achieved by encrypting it with the intended recipient's public key, so that only the private key from that same key pair can reverse the transformation.

In the exchange this question describes, the sender obtains the receiver's public key and uses it to encrypt the outgoing message. Because a private key can only undo encryption performed with its own matching public key, the secret capable of reversing this encryption belongs to the same key pair as the public key that was used -- that is, to the receiver -- and it is never handed to, or shared with, the sender or any other party.

• The role labelled sender only ever uses a public key -- an openly shared key -- to encrypt the outgoing message; that role never holds the private key from that same key pair, so attaching the decrypting secret to the encrypting role contradicts how this scheme works.

• This groups two different roles together as though both were given the same secret decrypting key. Only the key pair whose public half was used to encrypt has a matching private half in play here, and the role that performs the encryption step is not part of that pair -- so treating both roles as joint key-holders is incorrect.

• Exactly one key pair is in play for this encrypted message, and its private half sits with a single role. There is no scheme here in which several roles simultaneously hold that one secret half.

• One of the roles listed above is in fact the one whose key pair supplies the matching private half for this message, so ruling out every listed role overstates the case.

Note: in a fully two-way exchange each side may separately own its own key pair for its own inbound messages or for digital signing, but that does not change who holds the private key needed to decrypt this particular message -- the report's underlying observation is true in general, but does not apply to the single keyed role this question asks about.