The research work within the project-team is mostly devoted to the design and analysis of cryptographic algorithms, in the classical or in the quantum setting. It is especially motivated by the fact that the current situation of cryptography is rather fragile: many of the available symmetric and asymmetric primitives have been either threatened by recent progress in cryptanalysis or by the possible invention of a large quantum computer. Most of our work mixes fundamental aspects and practical aspects of information protection (cryptanalysis, design of algorithms, implementations). In particular we devise

work on practical aspects in cryptography, e.g. lightweight constructions and implementation, but also on more fundamental issues, either on discrete mathematics or on quantum information.

The current state-of-the-art asymmetric cryptography would become insecure in a post-quantum world, and the community is actively searching for alternatives. Symmetric cryptography, essential for enabling secure communications, used to seem much less affected at first sight: the biggest known threat was Grover's algorithm, which allows exhaustive key searches in the square root of the search space. Thus, it was believed that doubling key-lengths suffices to maintain an equivalent security in the post-quantum world. This conventional wisdom was contradicted by Kuwakado and Morii in 2012 when they proposed for the first time to use Simon's algorithm in symmetric cryptanalysis 81, proving the popular Even-Mansour construction to be insecure in a strong security model called the superposition model.

This model allows an attacker to query quantumly the block cipher. Simon's algorithm 83 contrarily to Grover's algorithm gives an exponential speedup and can therefore be devastating in this setting.

In the framework of our ERC QUASYModo, we studied in detail this algorithm and possible applications,
and we were able to show that Simon's algorithm applies to other schemes as well, such as for instance
to the CAESAR candidate AEZ 76. It also allows
to break some well-known modes of operation for MACs and authenticated encryption
and provides devastating quantum slide attacks 9. Other quantum algorithms turned out be useful in this model, such as for instance
Kuperberg's algorithm 80.
It allowed to break a
tweak 71
to counter the previous attack of 9 or to devise a quantum attack in the superposition model on the Poly1305 MAC primitive 74,
which is largely used and claimed to be quantumly secure.

All these results show that in symmetric (and asymmetric) cryptography, the impact of quantum computers goes
well beyond Grover's and Shor's algorithms and has to be studied carefully in order to understand if a given cryptographic
primitive is secure or not in a quantum world. To correctly evaluate the security of cryptographic primitives in the post-quantum
world, it is really desirable to elaborate a quantum cryptanalysis toolbox.
This is precisely the first objective of the ERC QUASYModo regarding symmetric cryptanalysis.
We plan in the coming years to continue to actively contribute to this toolbox. This goes together
with improving or finding new quantum algorithms for cryptanalysis, possibly adapted to some particular situations or scenarios that have not been studied before, like the

Symmetric techniques are widely used because they are the only ones that can achieve some major features such as high-speed or low-cost encryption, fast authentication, and efficient hashing. It is a very active research area which is stimulated by a pressing industrial demand for low-cost implementations. Even if the block cipher standard AES remains unbroken 20 years after its design, it clearly appears that it cannot serve as a Swiss Army knife in all environments. In particular an important challenge raised by several new applications is the design of symmetric encryption schemes with some additional properties compared to the AES, either in terms of implementation performance (low-cost hardware implementation, low latency, resistance against side-channel attacks...) or in terms of functionalities. The past decade has then been characterized by a multiplicity of new proposals and evaluating their security has become a primordial task which requires the attention of the community.

This proliferation of symmetric primitives has been amplified by public competitions, including the recent NIST lightweight standardization effort, which have encouraged innovative but unconventional constructions in order to answer the harsh implementation constraints. These promising but new designs need to be carefully analyzed since they may introduce unexpected weaknesses in the ciphers. Our research work captures this conflict for all families of symmetric ciphers. It includes new attacks and the search for new building blocks which ensure both a high resistance to known attacks and a low implementation cost. This work, which combines cryptanalysis and the theoretical study of discrete mathematical objects, is essential to progress in the formal analysis of the security of symmetric systems.

Our specificity, compared to most groups in the area, is that our research work tackles all aspects of the problem, from the practical ones (new attacks, concrete constructions of primitives and low-cost building-blocks) to the most theoretical ones (study of the algebraic structure of underlying mathematical objects, definition of optimal objects). We study these aspects not separately but as several sides of the same domain.

Current public-key cryptography is particularly threatened by quantum computers, since almost all cryptosystems used in practice rely on related number-theoretic security problems that can be easily solved on a quantum computer as shown by Shor in 1994. This very worrisome situation has prompted NIST to launch a standardization process in 2017 for quantum-resistant alternatives to those cryptosystems. This concerns all three major asymmetric primitives, namely public-key encryption schemes, key-exchange protocols and digital signatures. The NIST has made it clear that for each primitive there will be several selected candidates relying on different security assumptions. It publicly admits that the evaluation process for these post-quantum cryptosystems is significantly more complex than the evaluation of the SHA-3 and AES candidates for instance.

There were 69 (valid) submissions to this call in November 2017, with numerous lattice-based, code-based and multivariate-cryptography submissions and some submissions based either on hashing or on supersingular elliptic curve isogenies. In January 2019, 26 of these submissions were selected for the second round and 7 of them are code-based submissions. In July 2020, 15 schemes were selected as third round finalists/alternate candidates, 3 of them are code-based. NIST has anounced in 2021 that this call for postquantum primitives would be extended specifically for digital signatures based on techniques other than lattices. This new call should be released in the first quarter of 2022.

The research of the project-team in this field is focused on the design and cryptanalysis of cryptosystems making use of coding theory and we have proposed code-based candidates to the NIST call for the first two types of primitives, namely public-key encryption and key-exchange protocols and have two candidates among the finalists/alternate candidates. We are also preparing proposals of code-based signatures schemes for the call which is expected in 2022.

The field of quantum information and computation aims at exploiting the laws of quantum physics to manipulate information in radically novel ways. There are two main applications:

Our team deals with quantum coding theoretic issues related to building a large quantum computer and with quantum cryptography. If these two questions may seem at first sight quite distinct, they are in fact closely related in the sense that they both concern the protection of (quantum) information either against an adversary in the case of quantum cryptography or against the environment in the case of quantum error-correction. This connection is actually quite deep since an adversary in quantum cryptography is typically modeled by a party having access to the entire environment. The goals of both topics are then roughly to be able to measure how much information has leaked to the environment for cryptography and to devise mechanisms that prevent information from leaking to the environment in the context of error correction.

While quantum cryptography is already getting out of the labs, this is not yet the case of quantum computing, with large quantum computers capable of breaking RSA with Shor's algorithms maybe still decades away. The situation is evolving very quickly, however, notably thanks to massive public investments in the past couple of years and all the major software or hardware companies starting to develop their own quantum computers. One of the main obstacles towards building a quantum computer is the fragility of quantum information: any unwanted interaction with the environment gives rise to the phenomenon of decoherence which prevents any quantum speedup from occurring. In practice, all the hardware of the quantum computer is intrinsically faulty: the qubits themselves, the logical gates and the measurement devices. To address this issue, one must resort to quantum fault-tolerance techniques which in turn rely on the existence of good families of quantum error-correcting codes that can be decoded efficiently. Our expertise in this area lies in the study of a particularly important class of quantum codes called quantum low-density parity-check (LDPC) codes. The LDPC property, which is well-known in the classical context where it allows for very efficient decoding algorithms, is even more crucial in the quantum case since enforcing interactions between a large number of qubits is very challenging. Quantum LDPC codes solve this issue by requiring each qubit to only interact with a constant number of other qubits.

The research community is strongly involved in the
development and evolution of cryptographic standards. Many standards
are developed through open competitions (e.g. AES, SHA-3) where
multiple teams propose new designs, and a joint cryptanalysis effort
allows to select the most suitable proposals. The analysis of
established standards is also an important work, in order to depreciate
weak algorithms before they can be exploited. Several members of the
team have been involved in this type of effort and we plan to continue
this work to ensure that secure algorithms are widely available. We
believe that good cryptographic standards have a large socio-economic
impact, and we are active in proposing schemes to future
competitions, and in analyzing schemes proposed to current or future
competitions, as well as widely-used algorithms and standards.

At the moment, we are involved in the two standardization efforts run by NIST for post-quantum cryptography and lightweight cryptography. We have also uncovered potential backdoors in two algorithms from the Russian Federation (Streebog and Kuznyechik), and successfully presented the standardization of the latter by ISO. We have also implemented practical attacks against SHA-1 to speed-up its deprecation.

NIST post-quantum competition.

The NIST post-quantum competition1 aims at standardizing quantum-safe public-key
primitives. It is really about offering a credible quantum-safe alternative for the schemes based on number theory which are severely threatened by the advent of quantum computers.
It is expected to have a huge and long-term impact on all public-key cryptography. It has received 69 proposals in November 2017, among which five have been co-designed by the project-team.
Four of them have made it to the second round in January 2019. One of them was chosen in July 2020 for the third round and another one was chosen as an alternate third round finalist.
We have also broken two first round candidates Edon-K 82 and RankSign 79, and have devised a partial break of the RLCE encryption scheme 77. In 2020, we obtained a significant breakthrough in solving more efficiently the MinRank problem and the decoding problem in the rank metric 72, 73 by using algebraic techniques. This had several consequences: all second round rank metric candidates were dismissed from the third round (including our own candidate) and it was later found out that this algebraic algorithm could also be used to attack the third round multivariate finalist, namely Rainbow and the alternate third round finalist GeMSS.

NIST competition on lightweight symmetric encryption.

The NIST lightweight cryptography standardization process2 is an initiative to develop and standardize new authenticated encryption algorithms suitable for constrained devices. As explained in Subsection 3.2, there is a real need for new standards in lightweight cryptography, and the selected algorithms are expected to be widely deployed within the Internet of Things, as well as on more constrained devices such as contactless smart cards, or medical implants. The NIST received 56 submissions in February 2019, three of which have been co-designed by members of the team.

Monitoring Current Standards

While we are very involved in the design phase of new cryptographic standards (see above), we also monitor the algorithms that are already standardized. In practice, this work has two sides.

First, we work towards the deprecation of algorithms known to be unsage. Unfortunately, even when this fact is known in the academic community, standardizing bodies can be slow to implement the required changes to their standards. This prompted for example G. Leurent to implement even better attacks against SHA-1 to illustrate its very practical weakness, and L. Perrin and X. Bonnetain (then a COSMIQ member) to find simple arguments proving that a subfunction used by the current Russian standards was not generated randomly, despite the claims of its authors.

Second, it also means that we participate to the relevant ISO meetings discussing the standardization of cryptographic primitives (JC27/WG2), and that we follow the discussions of the IETF and IRTF on RFCs. We have also provided technical assistance to members of other standardizing bodies such as the ETSI.

Major academic and industrial efforts are currently underway to implement quantum key distribution at large scale by integrating this technology within existing telecommunication networks. Colossal investments have already taken place in China to develop a large network of several thousand kilometers secured by quantum cryptography, and there is little doubt that Europe will follow the same strategy, as testified by the current European projects CiViQ (in which we are involved), OpenQKD and the future initiative Euro-QCI (Quantum Communication Infrastructure). While the main objectives of these actions are to develop better systems at lower cost and are mainly engineering problems, it is crucial to note that the security of the quantum key distribution protocols to be deployed remains far from being completely understood. For instance, while the asymptotic regime of these protocols (where one assumes a perfect knowledge of the quantum channel for instance) has been thoroughly studied in the literature, it is not the case of the much more relevant finite-size regime accounting for various sources of statistical uncertainties for instance. Another issue is that compliance with the standards of the telecommunication industry requires much improved performances compared to the current state-of-the-art, and this can only be achieved by significantly tweaking the original protocols. It is therefore rather urgent to better understand whether these more efficient protocols remain as secure as the previous ones. Our work in this area is to build upon our own expertise in continuous-variable quantum key distribution, for which we have developed the most advanced security proofs, to give security proofs for the protocols used in this kind of quantum networks.

During the course of the COVID-19 pandemic, several digital tools were developped to help mitigating the pandemic. We have not been involved in the developpement of these tools, but we took an active role in analyzing them, and contributing to the political debate.

During the first wave of the COVID-19 pandemic, several efforts were initiated to develop smartphone applications intended to contribute to contact tracing. The core idea consists in using Bluetooth signal to estimate the distance and the duration of a contact between two app users.

Later, venue tracking was implemented in several countries. The core idea is to warn patrons when a public place is detected to be a cluster: patrons scan a QR-code with a random identifier when entering the venue, and a list of identifiers with known clusters is published daily.

In France Bluetooth tracing was implemented in the StopCovid application launched on June 2 2020, and renamed TousAntiCovid on October 22 2020. Venue tracking was added on June 9 2021.

At the end of 2020, discussions began in the European Union about vaccine passports and covid certificates, and the first guidelines from European institutions were published in January 2021. A Covid Certificate is a machine-readable document (usually in the form of a QR-code) containing health information with a cryptographic signature from a health autority.

Covid certificates started to be used in France on June 9 2021, and the European version was put in place on July 1st 2021.

Members of the COSMIQ team began to be involved in this topic in April 2020. As several contact tracing projects became public, an inter-disciplinary collaboration between researchers in cryptography, in security and in technology law, involving the COSMIQ, CARAMBA, PESTO project-teams and other academic institutions, was initiated in order to investigate the consequences of the deployment of such applications in terms of privacy and security. Indeed, a public (and often external) security analysis is always expected for applications dealing with sensitive data such as, in this instance, medical information and each user's social graph. As mentioned in the introduction of Inria's white book on cybersecurity, "the first step in cybersecurity is to identify threats and define a corresponding attacker model. [...] Since zero risk cannot exist, the early detection and mitigation of attacks is as important as the attempt to reduce the risk of successful attacks." Understanding the limits of a system is then necessary to improve its security and to decide whether it can be deployed without taking ill-considered risks, exactly as the side effects of a drug should be documented.

As political discussions and decisions were taking place, we contributed
to these debates by providing an easy to understand description of the
security pitfalls that are inherent to bluetooth-based contact tracing:
"le traçage anonyme, un bel oxymore" 75.
The analysis presented in 75 is, in most cases, independent of the
subtleties of the privacy-preserving mechanism, and in particular can be
applied to both so-called "centralized" and "decentralized" systems. As
a consequence, its authors also worked with researchers based in the UK
to provide an English translation https://tracing-risks.com/.

This work had a significant impact (the website received more than 100K unique visitors) and led to further contributions from researchers from the COSMIQ team.

During this second year of the COVID-19 pandemic, most conferences and workshops have been either cancelled or modified to be online events. Anne Canteaut played a significant role in enabling this transition as the program chair of Eurocrypt 2020 and Eurocrypt 2021. Eurocrypt 2021 was the first flagship conference in cryptography held in a hybrid format. The concomitance of remote talks and of in-person talks required to adapt the format of the conference, the lengths of the talks... This very first experience will motivate discussions on the future format of conferences in our area.

Our cryptanalysis results on SHA-1 10 and GEA 28 have helped convince users and industry to deprecate those obsolete standards. Publication of those attacks and discussion with industry has resulted in concrete actions to reduce usage of those ciphers.

Our project is also involved in two NIST competitions: the competition for lightweight cryptography and the competition for standardizing quantum safe cryptosystems. In the first competition, our team has still one candidate in the third round of the competition, while in the second competition we have one candidate that is a third round finalist and another one which is an alternate third round finalist. The outcome of these two competitions will have a strong impact since the standardized solutions will likely replace large parts of the world’s infrastructure underpinning secure global communication.

Anne Canteaut https://cyberwomenday-cefcys.com/en/

Thomas Debris-Alazard,
Cryptographie fondée sur les codes: nouvelles approches pour constructions et preuves; contribution en cryptanalyse, 78

Sorbonne Universités, UPMC University of Paris 6, 2019, https://www.societe-informatique-de-france.fr/2021/01/recherche-prix-de-these-gilles-kahn-laureats-2020/

We have kept on working on symmetric quantum cryptanalysis and generic quantum algorithms related to cryptanalysis, and in addition, started looking at some asymmetric cryptanalysis problems:

Our recent results in symmetric cryptography concern either the security analysis of existing primitives, or the design of new primitives. This second topic includes some work on the construction and properties of suitable building-blocks for these primitives, e.g. on the search of highly nonlinear functions.

Our work in this area is mainly focused on code-based cryptography, but some of our contributions, namely algebraic attacks, have applications in multivariate cryptography or in algebraic coding theory. Many contributions relate to the NIST call for postquantum primitives, either cryptanalysis or design.

We have also been organizing since 2015 a working group held every month or every two months on code-based cryptography that structures the French efforts on this topic: every meeting is attended by most of the groups working in France on this topic (project-team GRACE, University of Bordeaux, University of Limoges, University of Rennes and University of Rouen).

Our main contributions during the period are given below

Most of our work in quantum information deals with either quantum algorithms, quantum error correction or cryptography.

Orange Labs Caen (11/2019 -> 11/2022)
Funding for the supervision of Paul Frixon's PhD.

30 kEuros.

Bull-ATOS (07/2020 -> 06/2023)
Funding for the supervision of Maxime Rémaud's PhD.

60 kEuros.

Thalès (11/2020 -> 10/2023)
Funding for the supervision of Loïc Demange's PhD.

45 kEuros.

ANR SELECT (07/21→06/24)

Security Evaluation of Lightweight Encryption using new Cryptanalysis Techniques

ANR Program: AAP Générique 2020 (PRCI)

Partners: Inria COSMIQ, Nanyang Technological University (Singapour)

476 kEuros

In the last decades, we have seen a large deployment of smart devices and contact-less smart cards, with applications to the Internet of Things and smart cities. These devices have strong security requirements as they communicate sensitive data by radio, but they have very low resources available: constrained computing capabilities and limited energy. This led to security disasters with the use of weak home-made cryptography such as KeeLoq or MIFARE. More recently, the academic cryptography community has come up with dedicated lightweight designs such as PRESENT or Skinny, and the NIST is currently organizing a competition to select the next worldwide standards. The goal of this project is to perform a wide security evaluation of the designs submitted to the NIST competition, and of lightweight cryptographic algorithms in general. We will use latest cryptanalysis advances, but also propose new attacks; study classical attacks, but also physical ones (very powerful in such scenarios).

Informal International Partners

ERC QUASYModo

H2020 FET Flagship on Quantum Technologies - CiViQ

General purpose quantum computers must follow a fault-tolerant design to prevent ubiquitous decoherence processes from corrupting computations. All approaches to fault-tolerance demand extra physical hardware to perform a quantum computation. Kitaev's surface, or toric, code is a popular idea that has captured the hearts and minds of many hardware developers, and has given many people hope that fault-tolerant quantum computation is a realistic prospect. Major industrial hardware developers include Google, IBM, and Intel. They are all currently working toward a fault-tolerant architecture based on the surface code. Unfortunately, however, detailed resource analysis points towards substantial hardware requirements using this approach, possibly millions of qubits for commercial applications. Therefore, improvements to fault-tolerant designs are a pressing near-future issue. This is particularly crucial since sufficient time is required for hardware developers to react and adjust course accordingly.

This consortium will initiate a European co-ordinated approach to designing a new generation of codes and protocols for fault-tolerant quantum computation. The ultimate goal is the development of high-performance architectures for quantum computers that offer significant reductions in hardware requirements; hence accelerating the transition of quantum computing from academia to industry. Key directions developed to achieve these improvements include: the economies of scale offered by large blocks of logical qubits in high-rate codes; and the exploitation of continuous-variable degrees of freedom.

The project further aims to build a European community addressing these architectural issues, so that a productive feedback cycle between theory and experiment can continue beyond the lifetime of the project itself. Practical protocols and recipes resulting from this project are anticipated to become part of the standard arsenal for building scalable quantum information processors.

MCCL – Modular Code Cryptanalysis Library

Collaboration between CWI and Inria whose purpose is to improve the state of the art of the implementation of ISD (Information Set Decoding). In particular by solving new decoding challenges. This intiative is a follow-up of the July 2021 Inria-CWI workshop. The first meeting took place in Paris in November 2021 and gathered 12 people from both institutions.

ANR DEREC (10/16→03/22)

Relativistic cryptography

ANR Program: jeunes chercheurs

244 kEuros

The goal of project DEREC is to demonstrate the feasibility of guaranteeing the security of some cryptographic protocols using the relativistic paradigm, which states that information propagation is limited by the speed of light. We plan to study some two party primitives such as bit commitment and their security against classical and quantum adversaries in this model. We then plan to the integration of those primitives into larger cryptosystems. Finally, we plan on performing a demonstration of those systems in real life conditions.

ANR CBCRYPT (10/17→03/22)

Code-based cryptography

ANR Program: AAP Générique 2017

Partners: Inria COSMIQ (coordinator), XLIM, Univ. Rouen, Univ. Bordeaux.

197 kEuros

The goal of CBCRYPT is to propose code-based candidates to the NIST call aiming at standardizing public-key primitives which resist to quantum attacks. These proposals are based either on code-based schemes relying on the usual Hamming metric or on the rank metric. The project does not deal solely with the NIST call. We also develop some other code-based solutions: these are either primitives that are not mature enough to be proposed in the first NIST call or whose functionalities are not covered by the NIST call, such as identity-based encryption, broadcast encryption, attribute based encryption or functional encryption. A third goal of this project is of a more fundamental nature: namely to lay firm foundations for code-based cryptography by developing thorough and rigorous security proofs together with a set of algorithmic tools for assessing the security of code-based cryptography.

ANR quBIC (10/17→03/22)

Quantum Banknotes and Information-Theoretic Credit Cards

ANR Program: AAP Générique 2017

Partners: Univ. Paris-Diderot (coordinator), Inria COSMIQ, UPMC (LIP6), CNRS (Laboratoire Kastler Brossel)

87 kEuros

For a quantum-safe future, classical security systems as well as quantum protocols that guarantee security against all adversaries must be deployed. Here, we will study and implement one of the most promising quantum applications, namely unforgeable quantum money. A money scheme enables a secure transaction between a client, a vendor and a bank via the use of a credit card or via the use of banknotes, with maximal security guarantees. Our objectives are to perform a theoretical analysis of quantum money schemes, in realistic conditions and for encodings in both discrete and continuous variables, and to demonstrate experimentally these protocols using state-of-the-art quantum memories and integrated detection devices.

ANR SWAP (02/22→01/26)

Sboxes for Symmetric-Key Primitives

ANR Program: AAP Générique 2021

Partners: UVSQ (coordinateur), Inria COSMIQ, ANSSI, CryptoExperts, Univ. of Rouen, Univ. of Toulon.

172 kEuros

Sboxes are small nonlinear functions that are crucial components of most symmetric-key designs and their properties are highly related to the security of the overall construction. The development of new attacks has given rise to many Sbox design criteria. However, the emerge of new contexts, applications and environments requires the development of new design criteria and strategies. The SWAP project aims first at investigating such criteria for emerging use cases like whitebox cryptography, fully homomorphic encryption and side-channel resistance. Then, we wish for analyzing the impact of these particular designs on cryptanalysis and see how the use of Sboxes with some special mathematical structures can accelerate some known attacks or introduce new ones. Finally, we aim at studying Sboxes from a mathematical point of view and provide new directions to the Big APN problem, an old conjecture on the existence of a particular type of optimal permutations.

DIM SIRTEQ The SIRTEQ project labeled Major Interest Domain (DIM) is funded by the Ile-de-France Region. SIRTEQ brings together the largest European concentration of academic teams in the field of quantum technologies. Its main objective is to promote an excellent academic research in the field of quantum technologies in Ile de France, taking into account the actual current societal challenges and the importance of the transfer of knowledge and technologies.

We are involved in this project in the quantum communications (quantum cryptography) and quantum computation (quantum error codes, quantum cryptanalysis) themes.

Organization of the event “Rendez-vous des Jeunes Mathématiciennes et Informaticiennes” at Inria Paris (November 2-3) by C. Bouvier and A. Denys, a 2-day camp for 24 high-school girls interested in mathematics and computer science. J. Baudrin and C. Pernot conducted sessions there.

C. Pernot gave a talk in the event “Rendez-vous des Jeunes Mathématiciennes et Informaticiennes” at ENS on November 28.