This paper says in the appendix,

At a high-level, malicious privacy is defined similar to malicious security, but with a modified ideal functionality that is corruptible, meaning that the adversary can control what it outputs.

So does malicious privacy refer to malicious security without ensuring correctness? Then is malicious security stronger than malicious privacy?


1 Answer 1


Yes to both. I am not sure the name is extremely standard, but it is relatively common to view security as privacy+correctness. Similar notions have appeared in other works under different names (for example, I called it half-malicious security in this paper... But honestly, "malicious privacy" is a much better name).

  • $\begingroup$ Thank you for your answer. But I saw in your answer to this question that semi-malicious only destroys security and does not learn any private information. So is there not much difference between semi-malicous and malicious-privacy? $\endgroup$ Commented Nov 13, 2023 at 13:34
  • $\begingroup$ Damn, I never edited my answer to this question. If you read the comment section under my answer, you will see that pscholl pointed out that I confused "semi-malicious security" (which lets the adversary choose the random coins of the corrupted parties, but they play honestly otherwise) with "half-malicious security" (which is malicious privacy). After skimming through the literature, it seemed the common use of the term "semi-malicious" is that of pscholl, not the one I described in my answer. $\endgroup$ Commented Nov 13, 2023 at 13:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.