Publications

Can we give account-based cryptocurrencies anonymity among all accounts without making transaction size grow with the anonymity set?

We show full privacy is possible using a compact fully homomorphic encryption based transaction that correctly updates every account obliviously.

Can we compile semi-honest oblivious transfer into maliciously secure oblivious transfer in the minimum number of rounds?

We show the first such compiler in the plain model using correlation-intractable hashing to collapse a general cut-and-choose protocol.

Can we privately deliver and retrieve signals at million-user scale without revealing the sender or recipient?

We show orders-of-magnitude improvements using a trusted execution environment and oblivious RAM.

Can we securely extend STIR/SHAKEN to legacy telephone networks without broadcasting sensitive call metadata to every participating provider?

We show this is possible using Sidecar, a distributed privacy-preserving out-of-band signaling system with scalable protocols and tunable decentralization.

Can we retrieve every message for a recipient without linking repeated requests or limiting the message count?

We show an unrestricted, high-efficiency oblivious message retrieval protocol using two semi-honest, non-colluding servers and efficient cryptographic building blocks.

Can we trace abusive telephone calls in seconds without exposing call parties or carrier business data?

We show automated, privacy-preserving traceback using witness encryption, oblivious pseudorandom functions, and group signatures.

Can we compare what privacy today’s blockchain transaction systems actually provide?

We show how current systems trade privacy, scalability, usability, and compliance using a taxonomy of confidentiality, anonymity, unlinkability, and cryptographic techniques.

Can we release a cryptocurrency payment only when a threshold of oracles confirms a real-world event, without requiring oracle coordination?

We show secure conditional payments using verifiable witness encryption based on threshold signatures.

Can we turn two-round privacy-only oblivious transfer into maliciously secure, simulation-based oblivious transfer without setup or new assumptions?

We show a round-optimal four-round transformation using only black-box access to the original oblivious transfer and information-theoretic techniques.

Can buyers find matching suppliers on a decentralized marketplace without revealing their identity or sensitive requirements?

We show private interactive matching using a programmable blockchain, anonymous group signatures, and public-key encryption.

Can a recipient learn that a public bulletin board contains a message for them without revealing who they are or scanning the whole board?

We show full-private, constant-work signaling using either a trusted-execution server or two non-colluding servers with garbled circuits.

Can a phone prove it is not stolen without revealing its permanent equipment identifier?

We show anonymous blocklist non-membership using RSA accumulators and zero-knowledge proofs.

Can anonymous broadcast alone hide which stakeholder proposed a proof-of-stake block?

We show that realistic network delays can reveal the proposer using formal liveness–anonymity lower bounds and tagging attacks.

Can proof-of-stake protocols select eligible participants without revealing their identity or stake?

We show a modular anonymous-selection functionality using an instantiation based on Algorand’s selection function.