In abuse reporting, an established albeit discouraged practice
consists of redacting reported messages. In a nutshell, users may
have the ability to signal some received messages as being abusive.
Their mailbox providers should then wrap those messages according to
abuse reporting format (ARF, RFC 5965) and send them back to the
senders if they had subscribed to the relevant feedback loop. See
http://en.wikipedia.org/wiki/Feedback_loop_%28email%29
Redaction conceals the recipient, rather than the sender. In this
case, authors should markup the parts of the messages that contain
sensible data, in case the recipients report those messages as spam,
possibly inadvertently. A document describing how to carry out
redaction is now in LC, so it can hardly mention privicons. However,
privicons could consider this possible use. The redaction doc is at
http://datatracker.ietf.org/doc/draft-ietf-marf-redaction/
For casual readers, privicons are defined here:
http://tools.ietf.org/html/draft-koenig-privicons