Deniable Authentication and Non-Transferable Signatures

This direction is merged into another direction: deniable authentication. Its last version stays readable and citable here.

Version 1 of 129 papers45 conceptsfrozen 2026-10-04

In traditional digital signatures, authenticity and transferability are inextricably linked. If Alice signs a message, Bob can verify it came from her—and so can a judge, a newspaper, or anyone else Bob shows it to. This "public verifiability" is a feature for contracts but a bug for privacy.

Research into deniable authentication and non-transferable signatures seeks to decouple these properties. The goal is to allow Alice to convince Bob of her identity or the truth of a statement while ensuring that Bob cannot convince anyone else.

The Power to Lie: Non-Transferability through Forgery

The core intuition behind non-transferable signatures is a clever reversal of cryptographic logic: to make a proof deniable, we must give the recipient the technical means to forge it.

In Universal Designated-Verifier Signatures (UDVS), Ron Steinfeld, Laurence Bull, and Josef Pieprzyk describe a system where a signature is "designated" to a specific verifier. This verifier uses their own private key to process the signature. Crucially, the system is designed so that the verifier could have used that same private key to produce a bit-for-bit identical signature from scratch.

Because the verifier possesses the "trapdoor" required to simulate the signature, any third party looking at the result faces an evidentiary stalemate. They cannot know if Alice signed the message or if Bob forged it. As Xinyi Huang and colleagues note in their work on Short Designated Verifier Signatures, this symmetry—often built on Diffie-Hellman shared secrets—ensures that the proof is only meaningful to the intended recipient.

The Disjunctive Proof: "A or B"

To achieve this in more complex settings, researchers use disjunctive (OR) proofs. In this framework, a prover demonstrates knowledge of one of two things:

  1. The actual witness for the statement (e.g., Alice’s private key).
  2. A "trapdoor" secret held by the verifier.

In On the Possibility of One-Message Weak Zero-Knowledge, Boaz Barak and Rafael Pass use this to bypass impossibility results in the "plain model." By allowing a simulator to run in quasi-polynomial time, the simulator can "find" the trapdoor and produce a proof that is indistinguishable from one created by an honest prover.

This mechanism is also central to Concurrent Non-Malleable Commitments. Giuseppe Persiano, Rafail Ostrovsky, and Ivan Visconti show that by proving a disjunction between the committed value and a receiver-specific trapdoor, a protocol can remain secure even when an attacker tries to intercept and modify messages in transit.

Statistical Indistinguishability and Signer Ambiguity

For deniability to hold, the "fake" signature created by the verifier must be mathematically identical to the "real" signature created by the signer. If there is any statistical difference, an external observer can tell them apart, destroying the sender's privacy.

Maryam Rajabzadeh Asaar and Mohammad Reza Aref highlight this in their analysis of Code-based Strong Designated Verifier Signatures. They argue that signer ambiguity is lost if the mathematical structure of a simulated signature differs from an authentic one. True deniability requires "perfect ambiguity," where an observer has exactly a 50% chance of identifying the true signer, as explored in their work on Concurrent Signatures.

Deniability in Practice: From IKE to E-Health

These theoretical constructs have direct applications in securing communications:

  • Key Exchange: Andrew C Yao and colleagues proposed Deniable Internet Key-Exchange (IKE) by replacing digital signatures with Message Authentication Codes (MACs). Since both parties share the MAC key, either could have generated the authentication tag, leaving no non-repudiable trail.
  • E-Health: In Formal Analysis of an E-Health Protocol, Naipeng Dong and Hugo Jonker use "deceptive process simulation" to ensure receipt-freeness. This ensures that even if a doctor is coerced, they cannot prove what they prescribed to a third party.
  • Group Communication: Ivan Damgård and Claudio Orlandi have extended these ideas to Multi-Designated Verifier Signatures, allowing a message to be authentic to a whole group while remaining deniable to the outside world.

Resolving Disputes: Chameleon Hashing

A potential downside of deniability is that a malicious recipient could forge a signature and then falsely claim the sender actually signed it. To counter this, Xiaofeng Chen and others developed Key-Exposure Free Chameleon Hashing.

These schemes allow for selective convertibility: the original signer can provide a proof that a specific signature is a forgery without revealing their private key. This creates a "controlled" deniability where privacy is protected until a dispute arises, at which point the signer can choose to transform a private signature into a publicly verifiable one.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_deniable_authentication_and_non_transfer_a3ce738c,
  title = {Deniable Authentication and Non-Transferable Signatures},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/deniable-authentication-and-non-transferable-signatures/rdr_a3ce738c5ac24520beb1466262a3682c},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_deniable_authentication_and_non_transfer_a3ce738c_v1,
  title = {Deniable Authentication and Non-Transferable Signatures},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/deniable-authentication-and-non-transferable-signatures/rdr_a3ce738c5ac24520beb1466262a3682c/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}