Email | Twitter | Github | Linkedin
I am a PostDoc member of the Aarhus Crypto Group hosted by Ivan Damgård and Claudio Orlandi.
My PhD at the Technical University of Denmark (DTU) was awarded the DTU Compute fellowship. I hold a BSc from UCLA with distinction.
Prior to my PhD in computer science, I designed flight hardware for Mars missions at JPL NASA in Pasadena, advised global technology clients at the Boston Consulting Group, co-founded data-science startups in Switzerland and contributed to open-source libraries implementing the original Bitcoin protocol.
I am native in English and Swiss/German, advanced in Mandarin Chinese and intermediate in French.

Research Interests

Google Scholar / DBLP

I am interested in Privacy-Enhancing Technologies and Cryptographic Protocols for all domains.

During my PhD, I studied a class of smart contract applications called Decentralized Finance and characterized their economic security properties with formal verification techniques, identifying (front-running) vulnerabilities due to a lack of privacy. In response, my collaborators and I deployed advanced cryptography to design novel protocols which deliver input fairness in the permissionless setting and differentially private smart contracts enabled by secure multi-party computation (MPC).

Academic Service

PC: Program Committee , SR: Subreviewer
Cryptography & Security
ACM CCS'24, PC
Financial Crypto'24, PC
DeFi'24 (FC'24), PC
DeFi'23 (CCS'23), PC
CBT'23 (ESORICS'23), PC
DeFi'23 (FC'23), PC
DeFi'22 (CCS'22), PC
CBT'22 (ESORICS'22), PC
DeFi'22 (FC'22), PC

Crypto'24, SR
Eurocrypt'24, SR
CANS'24, SR
Eurocrypt'22, SR
Asiacrypt'22, SR
ACM CCS'22, SR
Formal Methods
WRLA'22, SR
FM'21, SR
TACAS'21, SR
JLAMP, Vol. 121, SR

Research Manuscripts

Click each paper title below for abstract

Detecting Rogue Decryption in (Threshold) Encryption via Self-Incriminating Proofs (ePrint)
Keeping decrypting parties accountable in public key encryption is notoriously hard since the secret key owner can decrypt any arbitrary ciphertext. Threshold encryption aims to solve this issue by distributing the power to decrypt among a set of parties, who must interact via a decryption protocol. However, such parties can employ cryptographic tools such as Multiparty Computation (MPC) to decrypt arbitrary ciphertexts without being detected. We introduce the notion of (threshold) encryption with Self-Incriminating Proofs, where parties must produce a self-incriminating proof of decryption when decrypting every ciphertext. In the standard public key encryption case, the adversary could destroy these proofs, so we strengthen our notion to guarantee that the proofs are published when decryption succeeds. This creates a decryption audit trail, which is useful in scenarios where decryption power is held by a single trusted party (e.g., a Trusted Execution Environment) who must be kept accountable. In the threshold case, we ensure that at least one of the parties who execute the decryption protocol will learn a self-incriminating proof, even if they employ advanced tools such as MPC. The fact that a party learns the proof and may leak it at any moment functions as a deterrent for parties who do not wish to be identified as malicious decryptors (e.g., a commercial operator of a service based on threshold encryption). We investigate the (im)possibility and applications of our notions while providing matching constructions under appropriate assumptions. In the threshold case, we build on recent results on Individual Cryptography (CRYPTO 2023).
James Hsin-yu Chiang, Bernardo David, Tore Kasper Frederiksen, Arup Mondal, Esra Yeniaras
Under Submission.
Correlated-Output Differential Privacy and Applications to Dark Pools (ePrint)
In the classical setting of differential privacy, a privacy-preserving query is performed on a private database, after which the query result is released to the analyst; a differentially private query ensures that the presence of a single database entry is protected from the analyst’s view. In this work, we contribute the first definitional framework for differential privacy in the trusted curator setting; clients submit private inputs to the trusted curator, which then computes individual outputs privately returned to each client. The adversary is more powerful than the standard setting; it can corrupt up to n − 1 clients and subsequently decide inputs and learn outputs of corrupted parties. In this setting, the adversary also obtains leakage from the honest output that is correlated with a corrupted output. Standard differentially private mechanisms protect client inputs but do not mitigate output correlation leaking arbitrary client information, which can forfeit client privacy completely. We initiate the investigation of a novel notion of correlated output differential privacy to bound the leakage from output correlation in the trusted curator setting. We define the satisfaction of both standard and correlated-output differential privacy as round differential privacy and highlight the relevance of this novel privacy notion to all application domains in the trusted curator model.

We explore round differential privacy in traditional "dark pool" market venues, which promise privacy-preserving trade execution to mitigate front-running; privately submitted trade orders and trade execution are kept private by the trusted venue operator. We observe that dark pools satisfy neither classic nor correlated-output differential privacy; in markets with low trade activity, the adversary may trivially observe recurring, honest trading patterns, and anticipate and front-run future trades. In response, we present the first round differentially private market mechanisms that formally mitigate information leakage from all trading activity of a user. This is achieved with fuzzy order matching, inspired by the standard randomized response mechanism; however, this also introduces a liquidity mismatch as buy and sell orders are not guaranteed to execute pairwise, thereby weakening output correlation; this mismatch is compensated for by a round differentially private liquidity provider mechanism, which freezes a noisy amount of assets from the liquidity provider for the duration of a privacy epoch, but leaves trader balances unaffected. We propose oblivious algorithms for realizing our proposed market mechanisms with secure multi-party computation (MPC) and implement these in the Scale-Mamba Framework using Shamir Secret Sharing based MPC. We demonstrate practical, round differentially private trading with comparable throughput as prior work implementing (traditional) dark pool algorithms in MPC; our experiments demonstrate practicality for both traditional finance and decentralized finance settings.
James Hsin-yu Chiang, Bernardo David, Mariana Gama, Christian Janos Lebeda
Advances in Financial Technologies, Princeton University, 2023 (doi)
FairPoS: Input Fairness in Permissionless Consensus (ePrint)
In permissionless consensus, the ordering of transactions or inputs in each block is freely determined by an anonymously elected block leader. A rational block leader will choose an ordering of inputs that maximizes financial gain; the emergence of automatic market makers in decentralized finance enables the block leader to front-run honest trade orders by injecting its own inputs prior to and after honest trades. Front-running is rampant in decentralized finance and reduces the utility of the system by extracting financial value from honest trades and increasing demand for block-space. Current proposals to prevent input order attacks by encrypting user inputs are not permissionless, as they rely on small static committees to perform distributed key generation and threshold decryption. Such committees require party authentication, knowledge of the number of participating parties or do not permit player replaceability and are therefore not permissionless. Moreover, alternative solutions based on sequencing inputs in order of their arrival cannot prevent front-running in an unauthenticated peer-2-peer network where message arrival is adversarially controlled.

We present FairPoS, the first consensus protocol to achieve input fairness in the permissionless setting with security against adaptive adversaries in semi-synchronous networks. In FairPoS, the adversary cannot learn the plaintext of any client input before it is included in a block in the chain's common-prefix. Thus, input ordering attacks that depend on observing pending client inputs in the clear are no longer possible. In FairPoS, this is achieved via Delay Encryption (DeFeo et al., EUROCRYPT 2021), a recent cryptographic primitive related to time-lock puzzles, allowing all client inputs in a given round to be encrypted under a key that can only be extracted after enough time has elapsed. In contrast to alternative approaches, the key extraction task in delay encryption can, in principle, be performed by any party in the permissionless setting and requires no distribution of secret key material amongst authenticated parties. However, key extraction requires highly specialized hardware in practice. Thus, FairPoS requires resource-rich staking parties to insert extracted keys into blocks, enabling light-clients to decrypt past inputs and relieving parties who join the execution from decrypting all inputs in the entire chain history. Realizing this in proof-of-stake is non-trivial; naive application of key extraction to proof-of-stake can result in chain stalls lasting the entire key extraction period. We overcome this challenge with a novel key extraction protocol, which tolerates adversarial delays in block delivery intended to prevent key extraction from completing on schedule. Critically, this also enables the adoption of a new longest-extendable-chain rule which allows FairPoS to achieve the same guarantees as Ouroborous Praos against an adaptive adversary.
James Hsin-yu Chiang, Bernardo David, Ittay Eyal, Tiantian Gong
Advances in Financial Technologies, Princeton University, 2023 (doi)
SoK: Privacy-Enhancing Technologies in Finance (ePrint)
Recent years have seen the emergence of practical advanced cryptographic tools that not only protect data privacy and authenticity, but also allow for jointly processing data from different institutions without sacrificing privacy. The ability to do so has enabled implementations a number of traditional and decentralized financial applications that would have required sacrificing privacy or trusting a third party. The main catalyst of this revolution was the advent of decentralized cryptocurrencies that use public ledgers to register financial transactions, which must be verifiable by any third party, while keeping sensitive data private. Zero Knowledge (ZK) proofs rose to prominence as a solution to this challenge, allowing for the owner of sensitive data (e.g. the identities of users involved in an operation) to convince a third party verifier that a certain operation has been correctly executed without revealing said data. It quickly became clear that performing arbitrary computation on private data from multiple sources by means of secure Multiparty Computation (MPC) and related techniques allows for more powerful financial applications, also in traditional finance.

In this SoK, we categorize the main traditional and decentralized financial applications that can benefit from state-of-the-art Privacy-Enhancing Technologies (PETs) and identify design patterns commonly used when applying PETs in the context of these applications. In particular, we consider the following classes of applications: 1. Identity Management, KYC & AML; and 2. Markets & Settlement; 3. Legal; and 4. Digital Asset Custody. We examine how ZK proofs, MPC and related PETs have been used to tackle the main security challenges in each of these applications. Moreover, we provide an assessment of the technological readiness of each PET in the context of different financial applications according to the availability of: theoretical feasibility results, preliminary benchmarks (in scientific papers) or benchmarks achieving real-world performance (in commercially deployed solutions). Finally, we propose future applications of PETs as Fintech solutions to currently unsolved issues. While we systematize financial applications of PETs at large, we focus mainly on those applications that require privacy preserving computation on data from multiple parties.
Carsten Baum, James Hsin-yu Chiang, Bernardo David, Tore Kasper Frederiksen
Advances in Financial Technologies, Princeton University, 2023 (doi)
Eagle: Efficient Privacy Preserving Smart Contracts (ePrint)
The proliferation of Decentralised Finance (DeFi) and Decentralised Autonomous Organisations (DAO), which in current form are exposed to front-running of token transactions and proposal voting, demonstrate the need to shield user inputs and internal state from the parties executing smart contracts. In this work we present "Eagle", an efficient UC-secure protocol which efficiently realises a notion of privacy preserving smart contracts where both the amounts of tokens and the auxiliary data given as input to a contract are kept private from all parties but the one providing the input. Prior proposals realizing privacy preserving smart contracts on public, permissionless blockchains generally offer a limited contract functionality or require a trusted third party to manage private inputs and state. We achieve our results through a combination of secure multi-party computation (MPC) and zero-knowledge proofs on Pedersen commitments. Although other approaches leverage MPC in this setting, these incur impractical computational overheads by requiring the computation of cryptographic primitives within MPC.
Our solution achieves security without the computation of cryptographic primitives inside the MPC instance and only requires a constant amount of exponentiations per client input.
Carsten Baum, James Hsin-yu Chiang, Bernardo David, Tore Kasper Frederiksen
Financial Cryptography and Data Security, Bol, Croatia, 2023
SoK: Mitigation of Front-running in Decentralized Finance (ePrint)
Front-running is the malicious, and often illegal, act of both manipulating the order of pending trades and injecting additional trades to make a profit at the cost of other users. In decentralized finance (DeFi), front-running strategies exploit both public knowledge of user trades from transactions pending on the network and the miner's ability to determine the final transaction order. Given the financial loss and increased transaction load resulting from adversarial front-running in decentralized finance, novel cryptographic protocols have been proposed to mitigate such attacks in the permission-less blockchain setting. We systematize and discuss the state-of-the-art of front-running mitigation in decentralized finance, and illustrate remaining attacks and open challenges.
Carsten Baum, James Hsin-yu Chiang, Bernardo David, Tore Kasper Frederiksen, Lorenzo Gentile
Workshop on Decentralized Finance (FC'22), Grenada, 2022 (doi)
Maximizing Extractable Value from Automated Market Makers (arXiv)
Automated Market Makers (AMMs) are decentralized applications that allow users to exchange crypto-tokens without the need for a matching exchange order. AMMs are one of the most successful DeFi use cases: indeed, major AMM platforms process a daily volume of transactions worth USD billions. Despite their popularity, AMMs are well-known to suffer from transaction-ordering issues: adversaries can influence the ordering of user transactions, and possibly front-run them with their own, to extract value from AMMs, to the detriment of users. We devise an effective procedure to construct a strategy through which an adversary can maximize the value extracted from user transactions.
Massimo Bartoletti, James Hsin-yu Chiang, Alberto Lluch-Lafuente
Financial Cryptography and Data Security, Grenada, 2022 (doi)
Formal analysis of lending pools in decentralized finance (arXiv)
Decentralised Finance (DeFi) applications constitute an entire financial ecosystem deployed on blockchains. Such applications are based on complex protocols and incentive mechanisms whose financial safety is hard to determine. Besides, their adoption is rapidly growing, hence imperilling an increasingly higher amount of assets. Therefore, accurate formalisation and verification of DeFi applications is essential to assess their safety. We have developed a tool for the formal analysis of one of the most widespread DeFi applications: Lending Pools (LP). This was achieved by leveraging an existing formal model for LPs, the Maude verification environment and the MultiVeStA statistical analyser. The tool supports several analyses including reachability analysis, LTL model checking and statistical model checking. In this paper we show how the tool can be used to analyse several parameters of LPs that are fundamental to assess and predict their behaviour. In particular, we use statistical analysis to search for threshold and reward parameters that minimize the risk of unrecoverable loans.
Massimo Bartoletti, James Hsin-yu Chiang, Tommi Junttila, Alberto Lluch Lafuente, Massimiliano Mirelli, Andrea Vandin
International Symposium on Leveraging Applications of Formal Methods, 2022 (doi)
A theory of Automated Market Makers in DeFi (arXiv)
Automated market makers (AMMs) are one of the most prominent decentralized finance (DeFi) applications. They allow users to exchange units of different types of crypto-assets, without the need to find a counter-party. There are several implementations and models for AMMs, featuring a variety of sophisticated economic mechanisms. We present a theory of AMMs. The core of our theory is an abstract operational model of the interactions between users and AMMs, which can be instantiated with any desired economic design mechanism. We exploit our theory to formally prove a set of fundamental properties of AMMs, characterizing both structural and economic aspects. We do this by abstracting from the actual economic mechanisms used in implementations, by identifying sufficient conditions which ensure the relevant properties. Notably, we devise a general solution to the arbitrage problem, the main game-theoretic foundation behind the economic mechanisms of AMMs.
Massimo Bartoletti, James Hsin-yu Chiang, Alberto Lluch-Lafuente
Coordination Models and Languages, 2021 (doi)
Logical Methods in Computer Science, Vol 18, Issue 4, 2022 (doi)
SoK: Lending Pools in Decentralized Finance (arXiv)
Lending pools are decentralized applications which allow mutually untrusted users to lend and borrow crypto-assets. These applications feature complex, highly parametric incentive mechanisms to equilibrate the loan market. This complexity makes the behaviour of lending pools difficult to understand and to predict: indeed, ineffective incentives and attacks could potentially lead to emergent unwanted behaviours. Reasoning about lending pools is made even harder by the lack of executable models of their behaviour: to precisely understand how users interact with lending pools, eventually one has to inspect their implementations, where the incentive mechanisms are intertwined with low-level implementation details. Further, the variety of existing implementations makes it difficult to distill the common aspects of lending pools. We systematize the existing knowledge about lending pools, leveraging a new formal model of interactions with users, which reflects the archetypal features of mainstream implementations. This enables us to prove some general properties of lending pools, such as the correct handling of funds, and to precisely describe vulnerabilities and attacks. We also discuss the role of lending pools in the broader context of decentralized finance.
Massimo Bartoletti, James Hsin-yu Chiang, Alberto Lluch-Lafuente
Workshop on Trusted Smart Contracts (FC'21), 2021 (doi)
Bitcoin Trace-Net: Formal Contract Verification at Signing Time (arXiv)
Smart contracting protocols promise to regulate the transfer of cryptocurrency amongst participants in a trustless manner. A safe smart contract implementation should ensure that each participant can always append a contract transaction to the blockchain in order move the contract towards secure completion. To this goal, we propose Bitcoin Trace-Net, a contract verification framework which generates an executable symbolic model from the underlying contract implementation. A Trace-Net model consists of a Petri Net formalism enriched with a Dolev-Yao-like actor knowledge model. The explicit symbolic actor knowledge model supports the verification of contracts featuring cryptographic sub-protocols, which may not be observable on the blockchain. Trace-Net is sufficiently expressive to accurately model blockchain semantics such as the delay between a transaction broadcast and its subsequent confirmation, as well as adversarial blockchain reorganizations of finite depths, both of which can break smart contract safety. As an implementation level framework, Trace-Net can be instantiated at run-time to monitor and verify smart contract protocol executions.

Other

Full course materials from my course on Bitcoin programming.
I contributed to cryptoeconomics authored by Eric Voskuil.