david wong

Hey! I'm David, a security engineer at the Blockchain team of Facebook, previously a security consultant for the Cryptography Services of NCC Group. I'm also the author of the Real World Cryptography book. This is my blog about cryptography and security and other related topics that I find interesting.

Hardware Solutions To Highly-Adversarial Environments - Whitebox Crypto vs TPM vs TEE vs Secure Enclaves vs Secure Elements vs HSM vs CloudHSM vs KMS - Part 1 posted 3 days ago

If at some point you realize that doing cryptography means having to manage long-term keys, it means you’re standing in the world of key management.

Makes sense right?

In these lands, you are going to run into scenarios where attackers can be quite close to your applications. I call these highly-adversarial environments.

Imagine using your credit card on an ATM skimmer (a doodads that a thief can place on top of the card reader of an ATM in order to copy the content of your credit card, see picture bellow); downloading an application on your mobile phone that compromises the OS; hosting a web application in a colocated server shared with a malicious customer; managing highly-sensitive secrets in a data center that gets breached; and so on.


These scenarios suck, and are very counterintuitive to most cryptographers. This is because cryptography has come a long way since the historical “Alice wants to encrypt a message to Bob without Eve intercepting it”. Nowadays, it’s often more like “Alice wants to encrypt a message to Bob, but Alice is also Eve”.

The key here is that in these scenarios, there’s not much that can be done cryptographically (unless you believe in whitebox crypto) and hardware can go a long way to help.

OK, so now we have a whole world of new doohickeys to learn about, and there’s a lot of thingamabob believe me (hence the dense title). It can be quite confusing to learn about all of this, so here we go: my promise is that by the end of this blogpost series you’ll have a better understanding of what are all these different hardware solutions.

Keep in mind that none of these solutions are pure cryptographic solutions: they are all defense-in-depth (and sometimes dubious) solutions that serve to hide secrets and their associated sensitive cryptographic operations. They also all have a given cost, meaning that if a sophisticated attacker decides to break the bank, there’s not much we can do (besides raising the cost of an attack).

OK let's get started.


By definition obfuscation has nothing to do with security: it is the act of scrambling something so that it still work but is hard to understand. So for the laugh, let’s first mention whitebox cryptography which attempts to “cryptographically” obfuscate the key inside of an algorithm. That’s right, you have the source code of some AES-based encryption algorithm with a fixed key, and it encrypts and decrypts fine, but the key is mixed so well with the implementation that it is too confusing for anyone to extract the key from the algorithm. That's the theory. Unfortunately in practice, no published whitebox crypto algorithm has been found to be secure, and most commercial solutions are closed-source due to this fact (security through obscurity kinda works in the real world). Again, it’s all about raising the cost and making it harder for attackers.

All in all, whitebox crypto is a big industry that sells dubious products to businesses in need of DRM solutions. On the more serious side, there is a branch of cryptography called Indistinguishability obfuscation (iO) that attempts to do this cryptographically (so for realz). iO is a very theoretical, impractical, and so far not-really-proven field of research. We’ll see how that one goes.

timeline whitebox crypto

(Timeline of whitebox cryptography, taken from Matthieu Rivain’s slides)

Smart Cards

OK, whitebox crypto is not great, and worse: even if you can’t extract the key, you can still copy the program instead of trying to extract the key (and use it to do whatever cryptographic operation it features). It would be great if we could prevent people from copying secrets from sensitive devices though, or even prevent them from seeing what’s going on when the device performs cryptographic operations. A smart card is exactly this. It’s what you commonly find in credit cards, and is activated by either inserting it into, or using Near-field Communication (NFC) by getting the smart card close enough to, a payment terminal (also called Point of Sale or PoS terminal).

smart cards

Smart cards are pretty old, and started as a practical way to get everyone a pocket computer. Indeed, a smart cart embarks a CPU, memory (RAM, ROM and EEPROM), input/output, hardware random number generator (so called TRNGs), etc.) unlike the not-so-smart cards that only had data stored in them via a magnetic stripe (which in turn can be easily copied via the skimmers I talked about previously). Today, it seems like the same people all have a much more powerful computer in their pockets, so smart cards are probably going to die. (Rob Wood is pointing to me that more than a quarter of the US still doesn’t have a smart phone, so there’s still some time before this prophecy come to fruition.)

Smart cards mix a number of physical and logical techniques to prevent observation, extraction, and modification of its execution environment and some of its memory (where secrets are stored). But as I said earlier, it’s all about how much money you want an attacker to be spending, and there exist many techniques that attempt at breaking these cards:

  • Non-invasive attacks such as differential power analysis (DPA) analyze the power consumption of the smart card while it is doing cryptographic operations in order to extract the associated keys.
  • Semi-invasive attacks require access to the chip’s surface to mount attacks such as differential fault analysis (DFA) which use heat, lasers, and other techniques to modify the execution of a program running on the smart card in order to leak the key via cryptographic attacks (see my post on RSA signature fault attacks for an example).
  • Finally invasive silicon attacks can modify the circuitry in the silicon itself to alter its function and reveal secrets.

Secure Elements

Smart cards got really popular really fast, and it became obvious that having such a secure blackbox in other devices could be useful. The concept of a secure element was born: a tamper-resistant microcontroller that can be found in a pluggable form factor like UUICs (SIM cards required by carriers to access their 3G/4G/5G network) or directly bonded on chips and motherboards like the embedded SE (eSE) attached to an iPhone’s NFC chip. Really just a small separate piece of hardware meant to protect your secrets and their usage in cryptographic operations.

SEs are an evolution of the traditional chip that resides in smart cards, which have been adapted to suit the needs of an increasingly digitalized world, such as smartphones, tablets, set top boxes, wearables, connected cars, and other internet of things (IoT) devices. (GlobalPlatform)

Secure elements are a key concept to protect cryptographic operations in the Internet of Things (IoTs), a colloquial (and overloaded) term to refer to devices that can communicate with other devices (think smart cards in credit cards, SIM cards in phones, biometric data in passports, garage keys, smart home sensors, and so on).

Thus, you can see all of the solutions that will follow in this blogpost series as secure elements implemented in different form factors, using different techniques, and providing different level of defense-in-depth.

If you are required to use a secure element (to store credit card data for example), you also most likely have to get it certified. The main definition and standards around a secure element come from GlobalPlatform, but there exist more standards like Common Criteria (CC), NIST’s FIPS, EMV (for Europay, Mastercard, and Visa), and so on. If you’re in the market of buying secure microcontrollers, you will often see claims like “FIPS 140-2 certified” and “certified CC EAL 5+” next to it. Claims that can be obtained after spending some quality time, and a lot of money, with licensed certification labs.

Host Card Emulation (HCE)

It’s 2020, most people have a computer in their pocket: a smart phone. What’s the point of a credit card anymore? Well, not much, nowadays more and more payment terminals support contactless payment via the Near-field Communication (NFC) protocol, and more and more smartphones ship with an NFC chip that can potentially act as a credit card.

NFC for payment is specified as Card Emulation. Literally: it emulates a bank card. Banks allow you to do this only if you have a secure element.

Since Apple has full control over its hardware, it can easily add a secure element to its new iPhones to support payment, and this is what Apple did (with an embedded SE bonded to the NFC chip since the iPhone 6). iPhone users can register a bank card with the Apple wallet application, Apple can then obtain the card’s secrets from the issuing bank, and the card secrets can finally be stored in the eSE. The secure element communicates directly with the NFC chip and then to NFC readers, thus a compromise of the phone OS does not impact the secure element.

Google, on the other hand, had quite a hard time introducing payment to Android-based mobile phones due to phone vendors all doing different things. The saving technology for Google ended up being a cloud-based secure element called Host Card Emulation (HCE) introduced in 2013 in Android 4.4.

How does it work? Google stores your credit card information in a secure element in the cloud (instead of your phone), and only gives your phone access to short-lived single-use PAN (card numbers). (Note that some Android devices do have an eSE that can be used instead of HCE, and some SIM cards can also be used as secure elements for payment.) This concept of replacing sensitive long-term information with short-lived tokens is called tokenization. Sending a random card number that can be linked to your real one is great for privacy: merchants can’t track you as it’ll look like you’re always using a new card number. If your phone gets compromised, the attacker only gets access to a short-lived secret that can only be used for a single payment. Tokenization is a common concept in security: replace the sensitive data with some random stuff, and have a table secured somewhere safe that maps the random stuff to the real data.

Wikipedia has some cool diagram to show what’s going on whenever you pay with Android Pay or Apple Pay:


Although Apple theoretically doesn't have to use tokenization, since iPhones have secure elements that can store the real PAN, they do use it in order to gain more privacy (it's afterall their new bread and butter).

In part 2 of this blog series I’ll cover TEEs, HSMs, TPMs, and much more :) so tune back later.

(I would like to thank Rob Wood, Thomas Duboucher, and Lionel Rivière for answering my many questions!)

PS: I'm writing a book which will contain this and much more, check it out!


Coronavirus and cryptography posted 2 weeks ago

The coronavirus is shaking the world, on its multiple layers, and cryptography hasn't been spared.

The IACR announced on March 14th that multiple conferences were postponed:

FSE 2020, which was supposed to be held in Athens, Greece, during 22-26 March 2020, has been postponed to 8-12 November 2020.

PKC 2020, which was supposed to be held in Edinburgh, Scotland, during 4-7 May 2020, has been postponed.

EUROCRYPT 2020, which was supposed to be held in Zagreb, Croatia, during 10-14 May 2020, has been postponed.

While some others were not:

No changes have been made at this time to the schedule of CRYPTO 2020, CHES 2020, TCC 2020, and ASIACRYPT 2020, but we will continue to closely monitor the situation and will inform members if changes are needed.

While many workplaces (including mine) are moving to a WFH (work from home) model, will conferences follow?

It seems to be the case at least for Consensus 2020, a cryptocurrency conference organized by coindesk, which is moving to an online model:

Consensus 2020 will now be a completely virtual experience, where attendees from all over the world can participate online at no charge.

On a more dramatic note it seems like several participants of EthCC, which was held in Paris almost a week ago, have contracted the virus. A google spreadsheet has been circulating in order to self-report and figure out who else could have potentially contracted the virus. Even Vitalik Buterin is rumored to have had mild COVID-19 symptoms. Nobody is out of reach.

On a lighter note, my coworker Kostas presented on proofs of solvency at the lightning talks of Real World Crypto 2020. With his merkle tree-like construction he hopes to make governments accountable when they count the number of people who counted positive to the virus.

comment on this story

EdDSA, Ed25519, Ed25519-IETF, Ed25519ph, Ed25519ctx, HashEdDSA, PureEdDSA, WTF? posted 3 weeks ago

The Edwards-curve Digital Signature Algorithm (EdDSA)

You've heard of EdDSA right? The shiny and new signature scheme (well new, it's been here since 2008, wake up).

Since its inception, EdDSA has evolved quite a lot, and some amount of standardization process has happened to it. It's even doomed to be adopted by the NIST in FIPS 186-5!

First, some definition:

  • EdDSA stands for Edwards-curve Digital Signature Algorithm. As its name indicates, it is supposed to be used with twisted Edwards curves (a type of elliptic curve). Its name can be deceiving though, as it is not based on the Digital Signature Algorithm (DSA) but on Schnorr signatures!
  • Ed25519 is the name given to the algorithm combining EdDSA and the Edwards25519 curve (a curve somewhat equivalent to Curve25519 but discovered later, and much more performant).

EdDSA, Ed25519, and the more secure Ed448 are all specified in RFC 8032.

RFC 8032: Edwards-Curve Digital Signature Algorithm (EdDSA)

RFC 8032 takes some new direction from the original paper:

  • It specifies a malleability check during verification, which prevents ill-intentioned people to forge an additional valid signature from an existing signature of yours. Whenever someone talks about Ed25519-IETF, they probably mean "the algorithm with the malleability check".
  • It specifies a number of Ed25519 variants, which is the reason of this post.
  • Maybe some other stuff I'm missing.

To sign with Ed25519, the original algorithm defined in the paper, here is what you're supposed to do:

  1. compute the nonce as HASH(nonce_key || message)
  2. compute the commitment R = [nonce]G with G the generator of the group.
  3. compute the challenge as HASH(commitment || public_key || message)
  4. compute the proof S = nonce + challenge × signing_key
  5. the signature is (R, S)

where HASH is just the SHA-512 hash function.


At a high-level this is similar to Schnorr signatures, except for the following differences:

  • The nonce is generated deterministically (as opposed to probabilistically) using a fixed nonce_key (derived from your private key, and the message M. This is one of the cool feature of Ed25519: it prevents you from re-using the same nonce twice.
  • The challenge is computed not only with the commitment and the message to sign, but with the public key of the signer as well. Do you know why?

Important: notice that the message here does not need to be hashed before being passed to the algorithm, as it is already hashed as part of the algorithm.

Anyway, we still don't know WTF all the variants specified are.

PureEdDSA, ContextEdDSA and HashEdDSA

Here are the variants that the RFC actually specifies:

  • PureEdDSA, shortened as Ed25519 when coupled with Edwards25519.
  • HashEdDSA, shortened as Ed25519ph when coupled with Edwards25519 (and where ph stands for "prehash").
  • Something with no name we'll call ContextEdDSA, defined as Ed25519ctx when coupled with Edwards25519.

All three variants can share the same keys. They differ only in their signing and verification algorithms.

By the way Ed448 is a bit different, so from now on I'll focus on EdDSA with the Edwards25519 curve.

Ed25519 (or pureEd25519) is the algorithm I described in the previous section.


Ed25519ctx (or ContextEd25519) is pureEd25519 with some additional modification: the HASH(.) function used in the signing protocol I described above is re-defined as HASH(x) = SHA-512(some_encoding(flag, context) || x) where:

  • flag is set to 0
  • context is a context string (mandatory only for Ed25519ctx)

In other words, the two instances of hashing in the signing algorithm now include some prefix. (Intuitively, you can also see that these variants are totally incompatible with each other.)

Right out of the bat, you can see that ContextEd25519 big difference is just that it mandates some domain separation to Ed25519.

Ed25519ph (or HashEd25519), finally, builds on top of ContextEd25519 with the following modifications:

  • flag is set to 1
  • context is now optional, but advised
  • the message is replaced with a hash of the message (the specification says that the hash has to be SHA-512, but I'm guessing that it can be anything in reality)

OK. So the big difference now seems that we are doubly-hashing.


Why HashEdDSA and why double hashing?

First, pre-hashing sucks, this is because it kills the collision resistance of the signature algorithm. In PureEdDSA we assume that we take the original message and not a hash. (Although this is not always true, the caller of the function can do whatever they want.) Then a collision on the hash function wouldn't matter (to make you create a signature that validates two different messages) because you would have to find a collision on the nonce which is computed using a secret (the nonce key).

But if you pre-hash the message, then finding a collision there is enough to obtain a signature that validates two messages.

Thus, you should use PureEdDSA if possible. And use it correctly (pass it the correct message.)

Why is HashEdDSA a thing then?

The EdDSA for more curves paper which was the first to introduce the algorithm has this to say:

The main motivation for HashEdDSA is the following storage issue (which is irrelevant to most well-designed signature applications). Computing the PureEdDSA signature of M requires reading through M twice from a buffer as long as M, and therefore does not support a small-memory “InitUpdate-Final” interface for long messages. Every common hash function H0 supports a smallmemory “Init-Update-Final” interface for long messages, so H0 -EdDSA signing also supports a small-memory “Init-Update-Final” interface for long messages. Beware, however, that analogous streaming of verification for long messages means that verifiers pass along forged packets from attackers, so it is safest for protocol designers to split long messages into short messages to be signed; this splitting also eliminates the storage issue.

Why am I even looking at this rabbit hole?

Because I'm writing a book, and it'd be nice to explain what the hell is going on with Ed25519.


What's a key exchange? posted 3 weeks ago


I've been writing about cryptography for a book for a year now, and it has brought me some interesting challenges. One of them is that I constantly have to throw away what I've learned a long time ago, and imagine what it feels like not to know about a concept.

For example what are key exchanges?

The most intuitive explanation that I knew of (up until recently) was the one given by the wikipedia page on key exchanges. You might already know about it (unless you're reading this post to learn about key exchanges). It's a picture that involves paint. Take a look at it, but don't try to understand what is going on if you don't know about key exchanges yet. You can come back to it later.


I thought this was great. At least until I tried to explain key exchanges to my friends using this analogy. Nobody got it.


The other problem was that I couldn't use colors to explain anything in my book, as it'll be printed in black & white.

So I sat on the sad realization that I didn't have a great explanation for key exchanges, this for a number of months, and that until a more intuitive idea came to my mind.


The idea goes like this. Imagine that Alice and Bob wants to share a secret, but are afraid that someone is intercepting their communications. What they do is that they go to the store and buy the same bottle of generic soda.

Once home, they both start a random timer and shake their respective bottles until their timer end.


What they obtain are some shaked, pressurized, ready to gush out bottles of sodas. Each of the bottles will release a different amount of pressure.

After that, they swap bottles. Now Alice has the bottle of Bob, and Bob has Alice's bottle.


What do they do now? They restart their timers and shake the other person's bottle for the same amount of time.

Shake shake shake!

shake again

What do they finally obtain? Try to guess.

If I did my job correctly, then I gave you an intuition of how key exchanges work. Both Alice and Bob should now have two bottle of sodas that will release the same pressure once opened. And that's the secret!

And even if I steal the two bottles, I can't get a bottle that combines both bottles' pressure.

I recap the whole flow in the picture below:


Did you know about key exchanges before? Did you get it? Or did you think the painting example made more sense?

Please tell me in the comment!

This is probably what I'll include in my book as an introduction of what key exchanges are, unless I find a better way to explain it :)


Cryptographic Signatures, Surprising Pitfalls, and LetsEncrypt posted 4 weeks ago

The Let's Encrypt Accident

On August 11th, 2015, Andrew Ayer posted the following email to the IETF mailing list:

I recently reviewed draft-barnes-acme-04 and found vulnerabilities in the DNS, DVSNI, and Simple HTTP challenges that would allow an attacker to fraudulently complete these challenges.

(The author has since then written a more complete explanation of the attack.)

The draft-barnes-acme-04 mentioned by Andrew Ayer is a document specifying ACME, one of the protocols behind the Let's Encrypt Certificate Authority. The thing that your browser trust and that signs the public keys of websites you visit.

The attack was found merely 6 weeks before major browsers were supposed to ship with Let's Encrypt's public keys in their trust store. The draft has since become RFC 8555: Automatic Certificate Management Environment (ACME), mitigating the issues. Since then no cryptographic attacks are known on the protocol.

But how did we get there? What's the deal with signature schemes these days? and are all of our protocols doomed? This is what this blog post will answer.

Let's Encrypt Use Of Signatures

Let's Encrypt is a pretty big deal. Created in 2014, it is a certificate authority ran as a non-profit, and currently providing trust to ~200 million of websites. (You can read my article Let's Encrypt Overview to learn more about it.)

The key to Let's Encrypt's success are two folds:

  • It is free. Before Let's Encrypt most certificate authorities charged fees from webmasters who wanted to obtain certificates.
  • It is automated. If you follow their standardized protocol, you can request, renew and even revoke certificates via an API. Contrast that to other certificate authorities who did most processing manually, and took time to issue certificates.

If a webmaster wants her website example.com to provide a secure connection to her users (via HTTPS), she can request a certificate from Let's Encrypt, and after proving that she owns the domain example.com and getting her certificate issued, she will be able to use it to negotiate a secure connection with any browser trusting Let's Encrypt.

That's the theory.

In practice the flow is the following:

  1. Alice registers on Let's Encrypt with an RSA public key.
  2. Alice asks Let's Encrypt for a certificate for example.com.
  3. Let's Encrypt asks Alice to prove that she owns example.com, for this she has to sign some data and upload it to example.com/.well-known/acme-challenge/some_file.
  4. Once Alice has signed and uploaded the signature, she asks Let's Encrypt to go check it.
  5. Let's Encrypt checks if it can access the file on example.com, if it successfully downloaded the signature and the signature is valid then Let's Encrypt issues a certificate to Alice.

I recapitulate some of this flow in the following figure:

let's encrypt flow

Now, you might be wondering, what if Alice does not own example.com and manage to man-in-the-middle Let's Encrypt in step 5? That's a real issue that's been bothering me ever since Let's Encrypt launched, and turns out a team of researchers at Princeton demonstrated exactly this in Bamboozling Certificate Authorities with BGP:

We perform the first real-world demonstration of BGP attacks to obtain bogus certificates from top CAs in an ethical manner. To assess the vulnerability of the PKI, we collect a dataset of 1.8 million certificates and find that an adversary would be capable of gaining a bogus certificate for the vast majority of domains

The paper continues and proposes two solutions to sort of remediate, or at least reduce the risk of these attacks:

Finally, we propose and evaluate two countermeasures to secure the PKI: 1) CAs verifying domains from multiple vantage points to make it harder to launch a successful attack, and 2) a BGP monitoring system for CAs to detect suspicious BGP routes and delay certificate issuance to give network operators time to react to BGP attacks.

Recently Let's Encrypt implemented the first solution multi-perspective domain validation, which changes the way step 5 of the above flow is performed: now Let's Encrypt downloads the proof from example.com from multiple places.

How Did The Let's Encrypt Attack Worked

But let's get back to what I was talking about, the attack that Andrew Ayer found in 2015.

In it, Andrew proposes a way to gain control of a Let's Encrypt account that has already validated a domain (let's say example.com)

The attack goes like this:

  1. Alice registers and goes through the process of verifying her domain example.com by uploading some signature over some data on example.com/.well-known/acme-challenge/some_file. She then successfully manages to obtain a certificate from Let's Encrypt.
  2. Later, Eve signs up to Let's Encrypt with a new account and new RSA public key, and request to recover the example.com domain
  3. Let's Encrypt asks Eve to sign some new data, and upload it to example.com/.well-known/acme-challenge/some_file (note that the file is still lingering there from Alice's previous domain validation)
  4. Eve crafts a new malicious keypair, and updates her public key on Let's Encrypt. She then asks Let's Encrypt to check the signature
  5. Let's Encrypt obtains the signature file from example.com, the signature matches, Eve is granted ownership of the domain example.com.

I recapitulate the attack in the following figure:

attack let's encrypt

Wait what?

What happened there?

Key Substitution Attack With RSA

In the above attack Eve managed to create a valid public key that validates a given signature and message.

This is because, as Andrew Ayer wrote:

A digital signature does not uniquely identify a key or a message

If you remember how RSA works, this is actually not too hard to understand.


For a fixed signature and (PKCS#1 v1.5 padded) message, a public key (e, N) must satisfy the following equation to validate the signature:

\(\text{signature} = \text{message}^e \pmod{N}\)

One can easily craft a public key that will (most of the time) satisfy the equation:

  • \(e = 1\)
  • \(N = \text{signature} - \text{message}\)

You can easily verify that the validation works:

$$ \begin{align} &\text{signature} = \text{message}^e \pmod{N}\\ \iff&\text{signature} = \text{message} \pmod{\text{signature} - \text{message}}\\ \iff&\text{signature} - \text{message} = 0 \pmod{\text{signature} - \text{message}} \end{align} $$

By definition the last line is true.

Security of Cryptographic Signatures

Is this issue surprising?

It should be.

And if so why?

This is because of the gap that exists between the theoretical world and the applied world, between the security proofs and the implemented protocol.

Signatures in cryptography are usually analyzed with the EUF-CMA model, which stands for Existential Unforgeability under Adaptive Chosen Message Attack.

In this model YOU generated a key pair, and then I request YOU to sign a number of arbitrary messages. While I observe the signatures you produce, I win if I can at some point in time produce a valid signature over a message I hadn't requested.

Unfortunately, even though our modern signature schemes seem to pass the EUF-CMA test fine, they tend to exhibit some surprising properties.

Subtle Behaviors of Signature Schemes

The excellent paper Seems Legit: Automated Analysis of Subtle Attacks on Protocols that Use Signatures by Dennis Jackson, Cas Cremers, Katriel Cohn-Gordon, and Ralf Sasse attempts to list these surprising properties and the signature schemes affected by them (and then find a bunch of these in protocols using formal verification, it's a cool paper read it).

Seems Legit: Automated Analysis of Subtle Attacks on Protocols that Use Signatures

Let me briefly describe each properties:

Conservative Exclusive Ownership (CEO)/Destructive Exclusive Ownership (DEO). This refers to what Koblitz and Menezes used to call Duplicate Signature Key Selection (DSKS). In total honesty, I don't think any of these terms are self-explanatory. I find these attacks easier to remember if thought of as the following two variants:

  1. key substitution attacks (CEO), where a different keypair or public key is used to validate a given signature over a given message.
  2. message key substitution attacks (DEO), where a different keypair or public key is used to validate given signature over a new message.

To recap: the first attack fixes both the message and the signature, the second one only fixes the signature.

Malleability. Most signature schemes are malleable, meaning that if you give me a valid signature I can tamper with it so that it becomes a different but still valid signature. Note that if I'm the signer I can usually create different signatures for the same message, but here malleability refers to the fact that someone who has zero knowledge of the private key can also create a new valid signature for the same signed message. It is not clear if this has any impact on any real world protocol, eventhough the bitcoin MtGox exchange blamed their loss of funds on this one. From the paper Bitcoin Transaction Malleability and MtGox:

In February 2014 MtGox, once the largest Bitcoin exchange, closed and filed for bankruptcy claiming that attackers used malleability attacks to drain its accounts.

Note that a newer security model called SUF-CMA (for strong EUF-CMA) attempts to include this behavior in the security definition of signature schemes, and some recent standards (like RFC 8032 that specifies Ed25519) are mitigating malleability attacks on their signature schemes.

Re-signability. This one is simple to explain. To validate a signature over message you often don't need the message itself but its digest. This would allow anyone to re-sign the message with their own keys without knowing the message itself. How is this impactful in real world protocols? Not sure, but we never know.

Collidability. This is another not-so-clear if it'll bite you one day: some schemes allow you to craft signatures that will validate under several messages. Worse, Ed25519 as designed allows one to craft a public key and a signature that would validate any message with high probability. (This has been fixed in some implementations like libsodium.)

I recapitulate the substitution attacks in the diagram below:

key substitution attack

What to do with all of this information?

Well, for one signature schemes are definitely not broken, and you probably shouldn't worry if your use of them are mainstream.

But if you're designing cryptographic protocols, or if you're implementing something that's more complicated than the every day use of cryptography you might want to keep these in the back of your mind.

book Did you like this content? This is part of a book about how to apply modern cryptography in real world applications. Check it out!


Authentication What The Fuck: Part II posted last month

Writing about real world cryptography, it seems like what I end up writing a lot about is protocols and how they solve origin/identity authentication.

Don't get me wrong, confidentiality has interesting problems to (e.g. how to bring confidentiality to a blockchain), but authentication is most of what applied cryptography is about, for realz.

Do I need to convince you?

If you think about it, most protocols are about finding ways to provide authentication to different scenarios. And that's why they can get complicated!

I'll take my life for example, here is the authentication problems and solutions that I use:

  • insecure → one-side authenticated. Every day I use HTTPS, which uses the web public-key infrastructure (web PKI) to allow my browser to authenticate any websites on the web. It's a mess, but that's how you scale machine-to-machine authentication nowadays.
  • one-side authenticated → mutually-authenticated. Whenever I log into a website, over a secure HTTPS connection, this is what happens. A machine asks me to present some password (in clear, or oblivious via an asymmetric password-authenticated key exchange), or maybe a one-time password (via TOTP), or maybe I'll have to press my thumb on a yubikey (FIDO 2), or maybe I'll have to do a combination of several things (MFA). These are usually machine authenticating humans-type of flow.
  • insecure → mutually-authenticated. Whenever I talk to someone on Signal, or connect to a new WiFi, or pair a bluetooth device (like my phone with a car), I go from an insecure connection to a mutually-authenticated connection. There is a bit more nuance here, as sometimes I'll authenticate a machine (a WiFi access point for example) and sometimes I'll authenticate a human (end-to-end encryption). So different techniques work best depending on the type of peer you're trying to talk to.

In the end, I think these are the main three big categories of origin authentication. Can you think of a better classification?

comment on this story

What Are Short Authenticated Strings (SAS)? posted last month

There is a whole field of authentication in cryptography that is often under-discussed (at least in my opinion).

See, we often like to talk about how key exchanges can also be authenticated, by mean of public-key infrastructures (e.g. HTTPS) or by pre-exchanging secrets (e.g. my previous post on sPAKE), but we seldom talk about post-handshake authentication.

Post-handshake authentication is the idea that you can connect to something (often a hardware device) insecurely, and then "augment" the connection via some information provided in a special out-of-band channel.

But enough blabla, let me give you a real-world example: you link your phone with your car and are then asked to compare a few digits (this pairing method is called "numeric comparison" in the bluetooth spec). Here:

  • Out-of-band channel. you are in your car, looking at your screen, this is your out-of-band channel. It provides integrity (you know you can trust the numbers displayed on the screen) but do not necessarily provide confidentiality (someone could look at the screen through your window).
  • Short authenticated string (SAS). The same digits displayed on the car's screen and on your phone are the SAS! If they match you know that the connection is secure.

This SAS thing is extremely practical and usable, as it works without having to provision devices with long secrets, or having the user compare long strings of unintelligible characters.

How to do this? You're probably thinking "easy!". An indeed it seems like we could just do a key exchange, and then pass the output in some KDF to create this SAS.


This has been discussed long-and-large on the internet: with key exchange protocols like X25519 it doesn't work. The reason is that X25519 does not have contributory behavior: I can send you a public key that will lead to a predictable shared secret. In other words: your public key does not contribute (or very little) to the output of the algorithm.

The correct™ solution here is to give more than just the key exchange output to your KDF: give it your protocol transcript. All the messages sent and received. This puts any man-in-the-middle attempts in the protocol to a stop. (And the lesson is that you shouldn't naively customize a key exchange protocol, it can lead to real world failures.)

The next question is, what's more to SAS-based protocols? After all, Sylvain Pasini wrote a 300-page thesis on the subject. I'll answer that in my next post.

comment on this story

How symmetric password-authenticated key exchanges work (sPAKE) posted last month

have you heard of sPAKE (or bPAKE)?

a sPAKE is first and foremost a PAKE, which stands for Password-Authenticated Key Exchange. This simply means that authentication in the key exchange is provided via the knowledge of a password. The s (resp. b) in front means symmetric (resp. balanced). This indicates that both sides know the password.

Alice and Bob trying to use a sPAKE to authenticate a key exchange

Other PAKEs where only one side knows the password exist, these are called aPAKE for asymmetric (or augmented) PAKEs. Yes I know the nomenclature is a bit confusing :)

The most promising sPAKE scheme currently seems to be SPAKE2, which is in the process of being standardized here. There are other sPAKEs, like Dragonfly which is used in WPA3, but they don't seem to provide as strong properties as SPAKE2.

The trick to a symmetric PAKE is to use the password to blind the key exchange's ephemeral keypairs.

The first part of a sPAKE with SPAKE2

Note that we can't use the password as is, instead we:

  • Pass the password into a memory-hard hash function like Argon2 to obtain w. Can you guess why we do this? (leave a comment if you do!)
  • Convert it to a group element. To do this we simply consider w a scalar and do a scalar multiplication with a generator of our subgroup (M or N depending if you're the client or the server, can you guess why we use different generators?)

NOTE: If you know BLS or OPAQUE, you might be wondering why we don't use a "hash-to-curve" algorithm, this is because we don't need to obtain a group element with an unknown discrete logarithm in SPAKE2.

Once the blinded (with the password) public keys have been exchanged, both sides can compute a shared group element:

  • Alice computes K = h × alice_private_key × (S - w × N)
  • Bob computes K = h × bob_private_key × (T - w × M)

Spend a bit of your time to understand these equations. What happens is that both Alice and Bob first unblind the public key they've received, then perform a key exchange with it, then multiply it with the value h. What's this value h? The cofactor, or simply put: the other annoying subgroup.

Finally Alice and Bob hash the whole transcript, which is the concatenation of:

  • Alice's identity.
  • Bob's identity.
  • The message Bob sent S.
  • The message Alice sent T.
  • The shared group element K.
  • The hardened password w.

The hash of this transcript gives us two things:

  • A shared secret !
  • A key that is further expanded (via a KDF) to obtain two authentication keys.

The key derivation of sPAKE

These authentication keys sole purpose is to provide key confirmation in the last round-trip of messages. That is to say at this point, if we don't do anything, we don't know if either Alice or Bob truly managed to compute the shared secret.

Key confirmation is pretty simple, both sides just have to compute an authentication tag with one of the authentication key produced over the transcript.

The final protocol looks a bit dense, but you should be able to decipher it if you've read this far.

Whole SPAKE2 protocol


Authentication, What The Fuck? posted January 2020

Authentication is an overloaded word in cryptography.

In the context of cryptographic primitives like message authentication codes (MACs) and authenticated encryption with associated data (AEAD), authentication really refers to authenticity or integrity. And as the Cambridge dictionary says:

Authenticity. the quality of being real or true.
The poems are supposed to be by Sappho, but they are actually of doubtful authenticity.
The authenticity of her story is beyond doubt.

The proof is in the pudding. When talking about the security properties of primitives like MACs, cryptography talks about unforgeability, which does relate to authenticity.

So whenever you hear things like "is this payload authenticated with HMAC?", think authenticity, think integrity.

In the context of protocols though (e.g. TLS) authentication refers to identification: the concept of proving who you are.

So whenever you hear things like "Is the server authenticated?", think "identities are being proven".

This dual sense really annoys me, but in the end this ambiguity is encompassed in the definition of authentication:

the process or action of proving or showing something to be true, genuine, or valid.

Diego F. Aranha proposes a clever way to disambiguate the two:

  • origin/entity authentication. You're proving that an entity really is who they say they are.
  • message authentication. You're proving that a message is genuine.

Note that an argument against this distinction is the following: to authenticate a message, you need a key. This key comes from somewhere (it's your context, or your "who"). So when you authenticate a message, you are really authenticating the context. This falls short in scenarios where for example you trust the root hash of a merkle tree, which authenticates all of its leaves.

The bottom line is, authentication is about proving that something is what it is supposed to be. And that thing can be a person, or a message, or maybe even something else.

This is not all. In the security world people are confused with authorization vs authentication :)

comment on this story

Messaging Layer Security: A Few Thoughts posted January 2020

I've been following the Messaging Layer Security (MLS) standardization a bit. I really appreciate what the people are doing there, and what they are trying to solve. I think group messaging is currently a huge mess, as every application I have seen/audited seemed to invent a new way to implement group chat. A common standard and guidelines would greatly help.

MLS' goal is to provide a solution to end-to-end encryption for group chats. A solution that scales.

If you don't know how the MLS protocol works, I advise you to read Michael Rosenberg's blog post or to watch the Real World Crypto talk on the subject (might not be available at the moment).

Thinking about the standard, I have two questions:

  1. Does a group chat loses any notion of privacy/confidentiality after it gets too large? For example, if you are in a Hong Kong group trying to organize a protest and there are more than 1,000 people in the group, what are the odds that one of them is a cop?
  2. Would a group chat protocol targeting groups with small numbers of participant (let's say 50 at most) be able to provide better security insurances efficiently?

For example, here are two security properties (taken from SoK: Secure Messaging) that MLS does not provide:

Speaker Consistency: All participants agree on the sequence of messages sent by each participant.

This means that if Alice (who is part of a group chat with Bob and Eve) colludes with the server, she can send "I like cats" to Bob and "I like dogs" to Eve.

Global Transcript: All participants see all messages in the same order. Note that this implies speaker consistency

This means that if Alice sends the following messages:

  • you must decide
  • your path

a server could re-order these messages so that Bob would see them in the same order, but Eve would see:

  • your path
  • you must decide


I have the following open questions:

  • Are these attacks important to protect against?
  • Is there an efficient protocol to prevent these attacks for groups of reasonable size?
  • If we cannot prevent them, can we detect them and warm the users?
  • If we are willing to change the protocol when going from 2 participants to 3 participants, would be willing to change the protocol when going from N to N+1 participants (where N is the number of participants threshold where confidentiality/privacy fades away)?

A history of end-to-end encryption and the death of PGP posted January 2020

This is were everything starts, we now have an open peer-to-peer protocol that everyone on the internet can use to communicate.

  • 1991
    • The US government introduces the 1991 Senate Bill 266, which attempts to allow "the Government to obtain the plain text contents of voice, data, and other communications when appropriately authorized by law" from "providers of electronic communications services and manufacturers of electronic communications service equipment". The bill fails to pass into law.
    • Pretty Good Privacy (PGP) - released by Phil Zimmermann.
  • 1993 - The US Government launches a criminal investigation against Phil Zimmermann for sharing a cryptographic tool to the world (at the time crypto exporting laws are a thing).
  • 1995 - Zimmermann publishes PGP's source code in a book via MIT Press, dodging the criminal investigation by using the first ammendment's protection of books.

That's it, PGP is out there, people now have a weapon to fight government surveillance. As Zimmermann puts it:

PGP empowers people to take their privacy into their own hands. There's a growing social need for it. That's why I wrote it.

  • 1995 - The RSA Data Security company proposes S/MIME as an alternative to PGP.
  • 1996
  • 1997
    • GNU Privacy Guard (GPG) - version 0.0.0 released by Werner Koch.
    • PGP 5 is released.

      The original agreement between Viacrypt and the Zimmermann team had been that Viacrypt would have even-numbered versions and Zimmermann odd-numbered versions. Viacrypt, thus, created a new version (based on PGP 2) that they called PGP 4. To remove confusion about how it could be that PGP 3 was the successor to PGP 4, PGP 3 was renamed and released as PGP 5 in May 1997

  • 1997 - PGP Inc is acquired by Network Associates
  • 1998 - RFC 2440 - OpenPGP Message Format

    OpenPGP - This is a definition for security software that uses PGP 5.x as a basis.

  • 1999
    • GPG version 1.0 released
    • Extensible Messaging and Presence Protocol (XMPP) is developed by the open source community. XMPP is a federated chat protocol (users can run their own servers) that does not have end-to-end encryption and requires communications to be synchronous (both users have to be online).
  • 2002 - PGP Corporation is formed by ex-PGP members and the PGP license/assets are bought back from Network Associates
  • 2004 - Off-The-Record (OTR) is introduced by Nikita Borisov, Ian Avrum Goldberg, and Eric A. Brewer as an extension of the XMPP chat protocol in "Off-the-Record Communication, or, Why Not To Use PGP"

    We argue that [...] the encryption must provide perfect forward secrecy to protect from future compromises [...] the authentication mechanism must offer repudiation, so that the communications remain personal and unverifiable to third parties

We now have an interesting development: messaging (which is seen as a different way of communication for most people) is getting the same security treatment as email.

  • 2006 - GPG version 2.0 released
  • 2007 - RFC 4880 - OpenPGP Message Format
  • 2010 - Symantec purchases the rights for PGP for $300 million.
  • 2011 - Cryptocat is released.
  • 2013 - The TextSecure (now Signal) application is introduced, built on top of the TextSecure protocol with Axolotl (now the Signal protocol with the double ratchet) as an evolution of OTR and SCIMP. It provides asynchronous communication unlike other messaging protocols, closing the gap between messaging and email.
  • 2014

PGP becomes increasingly criticized, as Matt Green puts it in 2014:

It’s time for PGP to die.

Another unexpected development: security professionals are now giving up on encrypted emails, and are moving to secure messaging. Is messaging going to replace email, even though it feels like a different mean of communication?

Moxie's quotes are quite interesting:

In the 1990s, I was excited about the future, and I dreamed of a world where everyone would install GPG. Now I’m still excited about the future, but I dream of a world where I can uninstall it.

In addition to the design philosophy, the technology itself is also a product of that era. As Matthew Green has noted, “poking through an OpenPGP implementation is like visiting a museum of 1990s crypto.” The protocol reflects layers of cruft built up over the 20 years that it took for cryptography (and software engineering) to really come of age, and the fundamental architecture of PGP also leaves no room for now critical concepts like forward secrecy.

In 1997, at the dawn of the internet’s potential, the working hypothesis for privacy enhancing technology was simple: we’d develop really flexible power tools for ourselves, and then teach everyone to be like us. Everyone sending messages to each other would just need to understand the basic principles of cryptography. [...]

The GnuPG man page is over sixteen thousand words long; for comparison, the novel Fahrenheit 451 is only 40k words. [...]

Worse, it turns out that nobody else found all this stuff to be fascinating. Even though GPG has been around for almost 20 years, there are only ~50,000 keys in the “strong set,” and less than 4 million keys have ever been published to the SKS keyserver pool ever. By today’s standards, that’s a shockingly small user base for a month of activity, much less 20 years.

  • 2018
    • the first draft of Messaging Layer Security (MLS) is published, a standard for end-to-end encrypted group chat protocols.
    • EFAIL releases damaging vulnerabilities against most popular PGP and S/Mime implementations.

      In a nutshell, EFAIL abuses active content of HTML emails, for example externally loaded images or styles, to exfiltrate plaintext through requested URLs. To create these exfiltration channels, the attacker first needs access to the encrypted emails, for example, by eavesdropping on network traffic, compromising email accounts, email servers, backup systems or client computers. The emails could even have been collected years ago.

  • 2019 - Latacora - The PGP Problem

    Why do people keep telling me to use PGP? The answer is that they shouldn’t be telling you that, because PGP is bad and needs to go away.

EFAIL is the straw that broke the camel's back. PGP is officially dead.

  • 2019
    • Matrix is out of beta and working on making end-to-end encryption the default.
    • Moxie gives a controversial talk at CCC arguing that advancements in security, privacy, censorship resistance, etc. are incompatible with slow moving decentralized protocols. Today, most serious end-to-end encrypted messaging apps use the Signal protocol (Signal, Facebook Messenger, WhatsApp, Skype, etc.)
    • XMPP's response: Re: the ecosystem is moving
    • Matrix's response: On privacy versus freedom

did you like this? This will part of a book on cryptography! Check it out here.


Difference between shamir secret sharing (SSS) vs Multisig vs aggregated signatures (BLS) vs distributed key generation (dkg) vs threshold signatures posted December 2019

That title is a mouthful! But so is the field.

Let me introduce the problem: Alice owns a private key which can sign transactions. The problem is that she has a lot of money, and she is scared that someone will target her to steal all of her funds.

Cryptography offers some solutions to avoid this being a key management problem.

The first one is called Shamir Secret Sharing (SSS), which is simply about splitting the signing private key into n shares. Alice can then split the shares among her friends. When Alice wants to sign a transaction, she would then have to ask her friends to give her back the shares, that she can use to recreate the signing private key. Note that SSS has many many variants, for example VSSS allows participants to verify that malicious shares are not being used, and PSSS allows participants to proactively rotate their shares.


This is not great though, as there is a small timeframe in which Alice is the single point of failure again (the moment she holds all the shares).

A logical next step is to change the system, so that Alice cannot sign a transaction by herself. A multi-signature system (or multisig) would require n participants to sign the same transaction and send the n signatures to the system. This is much better, except for the fact that n signatures means that the transaction size increases linearly with the number of signers required.


We can do better: a multi-signature system with aggregated signatures. Signature schemes like BLS allow you to compress the n signatures in a single signature. Note that it is currently much slower than popular signature schemes like ECDSA and EdDSA, so there must be a trade off between speed and size.

We can do even better though!

So far one still has to maintain a set of n public keys so that a signature can be verified. Distributed Key Generation (DKG) allows a set of participant to collaborate on the construction of a key pair, and on signing operations. This is very similar to SSS, except that there is never a single point of failure. This makes DKG a Multi-Party Computation (MPC) algorithm.

The BLS signature scheme can also aggregate public keys into a single key that will verify their aggregated signatures, which allows the construction of a DKG scheme as well.

Interestingly, you can do this with schnorr signatures too! The following diagram explains a simplified version of the scheme:

schnorr dkg

Note two things:

  • All these schemes can be augmented to become threshold schemes: we don't need n signatures from the n signers anymore, but only a threshold m of n. (Having said that, when people talk about threshold signatures, they often mean the threshold version of DKG.) This way if someone loses their keys, or is on holiday, we can still sign.
  • Most of these schemes assume that all participants are honest and by default don't tolerate malicious participants. More complicated schemes made to tolerate malicious participants exist.

Unfortunately all of this is pretty new, and as an active field of study no standard has been decided on one algorithm so far.

That's the difference!

One last thing: there's been some recent ideas to use zero knowledge proofs (ZKP) to do what aggregated signatures do but for multiple messages (because all the previous solutions all signed the same message). The idea is to release a proof that you have verified all the signatures associated to a set of messages. If the zero knowledge proof is shorter than all the signatures, it did its job!

did you like this? This will part of a book on cryptography! Check it out here.

EDIT: thanks to Dowhile and bascule for pointing errors in the post.


Writing a book is hard posted October 2019

I am now half-way in the writing of my book (I wrote 8 chapters out of 16) and I am already exhausted. It doesn't help that I started writing right before accepting a new position for a very challenging (and interesting) project. But here I am, half-way there, and I think I'm onto something. I can't wait to get there and look at the finished project as a real paper book :)

To give you some insight into this process, let me share some thoughts.

Writing is hard. I have realized that I need at least a full day to write something. It does take time to get into the zone, and writing in the morning before work just doesn't work for me (and writing after work is even worse). As JP Aumasson put it (about his process of writing Serious Cryptography):

I quickly realized that I didn’t know everything about crypto. The book isn’t just a dump of my own knowledge, but rather the fruit of hours of research—sometimes a single page would take me hours of reading before writing a single word.

So when I don't have a full day ahead of me, I use my limited time to read articles and do research in topics that I don't fully understand. This is useful, and I make more progress during the week end once I have time to write.

Revising is hard. If writing a chapter takes some effort X, revising a chapter takes effort X^3 . After each chapter, several people at Manning, and in my circle, provide feedback. At the same time, I realize that there's much more I want to write about subject Y and I start pilling up articles and papers that I want to read before I revise the chapter. I end up spending a TON of effort revising and re-visiting chapters.

Getting feedback is hard. I am lucky, I know a lot of people with different levels of knowledge in cryptography. This is very useful when I want to test how different audiences read different chapters. Unfortunately people are good at providing good feedback, and bad at providing bad feedback. And only the bad feedback ends up being useful feedback. If you want to help, [the first chapters are free to read](https://www.manning.com/books/real-world-cryptography?a_aid=Realworldcrypto&a_bid=ad500e09 ) and I'm ready to buy you a beer for some constructive negative feedback.

Laying out a chapter is hard. Writing a blog is relatively easy. It's short, self-contained, and often something I've been thinking about for weeks, months, before I put it into writing. Writing a chapter for a book is more like writing a paper: you want it to be perfect. Knowing a lot about the subject makes this even more difficult: you know you can make something great and not achieving that would be disappointing. One strategy that I wish I would have more time to spend on is the following one:

  • create a presentation about the subject of a chapter
  • give the presentation and observe what diagrams need revisiting and what parts are hard for an audience to understand
  • after many iterations put the slides into writing

I'm convinced this is the right approach, but I am not sure how I could optimize for this. If you're in SF and wants me to give you a presentation on one of the chapter of the book, leave a comment here :)


Algorand's cryptographic sortition posted September 2019

There are several cryptocurrencies that are doing really interesting things, Algorand is one of them. Their breakthrough was to make a leader-based BFT algorithm work in a permissionless setting (and I believe they are the first ones who managed to do this). At the center of their system lies a cryptography sortition algorithm. It's quite interesting, so I made a video to explain it!

PS: I've been doing these videos for a while, and I still don't have a cool intro, so if you want to make me a cool intro please do :D


What's my birthday? posted September 2019

My colleague Mesut asked me if using random identifiers of 128-bit would be enough to avoid collisions.

I've been asked similar questions, and every time my answer goes something like this:

you need to calculate the number of outputs you need to generate in order to get good odds of finding collisions. If that number is impressively large, then it's fine.

The birthday bound is often used to calculate this. If you crypto, you must have heard of something like this:

with the SHA-256 hash function, you need to generate at least 2128 hashes in order to have more than 50% chance of finding collisions.

And you know that usually, you can just divide the exponent of your domain space by two to find out how much output you need to generate to reach such a collision.

Now, this figure is a bit deceiving when it comes to real world cryptography. This is because we probably don't want to define "OK, this is bad" as someone reaching the point of having 50% chance of finding a collision. Rather, we want to say:

someone reaching one in a billion chance (or something much lower) to find a collision would be bad.

In addition, what does it mean for us? How many identifiers are we going to generate per second? How much time are we willing to keep this thing secure?

To truly answer this question, one needs to plug in the correct numbers and play with the birthday bound formula. Since this is not the first time I had to do this, I thought to myself "why don't I create an app for this?" and voila.

birthday bound

You can play with it here.

Thanks to my tool, I can now answer Mesut's question:

If you generate one million identifiers per second, in 26 years you will have one in a billion chance to generate a collision. Is this enough? If this is not adversary-controlled, or it is rate-limited, you will probably not generate millions of identifiers per second though, but rather thousands, in this case it will take 265 centuries to get these odds.