Say I'm a researcher like Satoshi Nakamoto that wants to stay anonymous. I want to publish documents and sign them somehow. A few years, I want to reveal my identity and prove it was me who was Satoshi all along.

Is there a cryptography scheme that can be used to achieve this?

  • 1
    $\begingroup$ Maybe I'm missing something. Just digitally sign the document(s) (really, sign their hashes) and publish the signatures next to the docs. Revealing the public key years later will allow the public to verify the signature. $\endgroup$ – Dan Oct 22 '18 at 23:21
  • $\begingroup$ Do we trust Satoshi? $\endgroup$ – kelalaka Oct 22 '18 at 23:25

The solution is easy if you have a Trusted Authority(TA). Assuming that there is no TA.

A direct digital signature can be a solution. However, how we can trust that Satoshi is not faking person $x$ by stealing his keys or he killed the person $x$ to steal the keys.

A solution

by using biometrics to uniquely identify the original person.

Satoshi follows the steps;

  • produces a private-public keys $k_{priv_s}, k_{pub_s}$ for his digital signatures.

  • he appends his biometric data; his fingerprint $fp_s$, his unique DNA code $dna_s$, and his iris $iris_s$ data to the document $d$.

    $$author = d \| fp_s \| dna_s \| iris_s $$

  • $Sign(k_{priv_s},author)$ by using digital signature standard.

  • Adds the signature and $k_{pub_s}$ to the document before print.

  • Secretly publishes the document.

So that, when a person showed up with the signature key to reveal himself, the others have to look at the biometric data to match.

The entropies;

Possible values: O’Gorman,L., Comparing Passwords, Tokens, and Biometrics for User Authentication;

A biometric doesn’t have a fixed number of possible values. Theoretically, the keyspace of biometrics such as fingerprints is unlimited because if you could measure the continuous signal with infinite precision, no two would be the same. But, one could say the same for passwords, that if you allowed the password length to be unlimited, you’d also have an unlimited keyspace

Fingerprints; Young et. al, Entropy of Fingerprints;

The average number of minutiae present in each image was 28.02. Therefore, dividing the entropy of 55.02 bits by the average minutiae of 28.02, each minutia provides 1.963 bits of entropy.

  1. Note: Fingerprint copying; I've found only this, it is easy to detect it when there are observers.
  2. Note: I changed the key generation from biometrics. Some of them are not exactly reliable.

Claims to be the Satoshi

  • nChain chief scientist Craig S. Wright that he created Bitcoin. But it is debunked and the claims are called nonsense.
  • 2
    $\begingroup$ thumbs $\textit{can}$ be stolen... $\endgroup$ – Ken Goss Oct 23 '18 at 1:02
  • 2
    $\begingroup$ How much entropy do these biometric markers possess, even if we assume no one has any idea what their fingerprint/voice print/etc are like? Which it should be added isn't a realistic assumption in the modern world, many countries take biometric data for ID and when coming into/exiting the country. Biometric data makes a better user name as opposed to a password. $\endgroup$ – Ella Rose Oct 23 '18 at 1:09
  • $\begingroup$ Post edit, I want to up vote this again. $\endgroup$ – Ken Goss Oct 23 '18 at 15:38

Simultaneous with the paper publish two additional files 1) an encrypted version of the file which is encrypted with a large asymmetric key 2) a file containing the decryption key. Then at any time you can prove you are the one who published the lot of it by later publishing the key used for encryption.

  • $\begingroup$ keys can be stolen. $\endgroup$ – kelalaka Oct 22 '18 at 23:17
  • 3
    $\begingroup$ Anything can be stolen. This doesn't guard against thieves, but it can allow the author to prove authorship, which was the goal of the question. $\endgroup$ – Ken Goss Oct 22 '18 at 23:19
  • 2
    $\begingroup$ Why exactly should someone use an asymmetric component to encrypt something in an effort to provide authenticity, especially considering that confidentiality is not required? An asymmetric component can instead be used for a digital signature, which is explicitly designed to provide authenticity and has actually been studied for such a use case. $\endgroup$ – Ella Rose Oct 23 '18 at 0:10
  • $\begingroup$ @EllaRose Many digital signature schemes are built out of asymmetric cryptosystems. At the root of it, your answer and mine are very similar. $\endgroup$ – Ken Goss Oct 23 '18 at 0:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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