Deep learning models are vulnerable to privacy attacks due to their tendency to memorize individual training examples. Theoretically-sound defenses such as differential privacy can defend against this threat, but model performance often suffers. Empirical defenses may thwart existing attacks while maintaining model performance but do not offer any robust theoretical guarantees. In this paper, we explore a new strategy based on the concept of plausible deniability. We introduce a training algorithm called **P**lausibly **D**eniable **S**tochastic **G**radient **D**escent (PD-SGD). The core of this approach is a rejection sampling technique, which probabilistically prevents updating model parameters whenever a mini-batch cannot be plausibly denied. We provide theoretical results showing that PD-SGD effectively mitigates privacy leakage from individual data points. Experiments demonstrate the scalability of PD-SGD and the favorable privacy-utility trade-off it offers compared to existing defense methods.

In this work, we propose a new privacy-preserving membership query protocol that lets a centralized entity privately query datasets held by one or more other parties to check if they contain a given element. This protocol, based on elliptic curve-based ElGamal and oblivious key-value stores, ensures that those 'data-augmenting' parties only have to send their encrypted data to the centralized entity once, making the protocol particularly efficient when the centralized entity repeatedly queries the same sets of data. We apply this protocol to detect anomalies in cross-silo federations. Data anomalies across such cross-silo federations are challenging to detect because (1) the centralized entities have little knowledge of the actual users, (2) the data-augmenting entities do not have a global view of the system, and (3) privacy concerns and regulations prevent pooling all the data. Our protocol allows for anomaly detection even in strongly separated distributed systems while protecting users' privacy. Specifically, we propose a cross-silo federated architecture in which a centralized entity (the backbone) has labeled data to train a machine learning model for detecting anomalous instances. The other entities in the federation are data-augmenting clients (the user-facing entities) who collaborate with the centralized entity to extract feature values to improve the utility of the model. These feature values are computed using our privacy-preserving membership query protocol. The model can be trained with an off-the-shelf machine learning algorithm that provides differential privacy to prevent it from memorizing instances from the training data, thereby providing output privacy. However, it is not straightforward to also efficiently provide input privacy, which ensures that none of the entities in the federation ever see the data of other entities in an unencrypted form. We demonstrate the effectiveness of our approach in the financial domain, motivated by the PETs Prize Challenge, which is a collaborative effort between the US and UK governments to combat international fraudulent transactions. We show that the private queries significantly increase the precision and recall of the otherwise centralized system and argue that this improvement translates to other use cases as well.

Hot/cold wallet refers to a widely used paradigm to enhance the security level of cryptocurrency applications that was proposed on Bitcoin Improvement Proposal 32. In a nutshell, after performing an initial setup in which the hot wallet receives partial information of the cold wallet in order to hierarchically generate (transaction receiving) addresses, the cold wallet stays offline, whereas the hot wallet is kept online. The initial transferred information enables the hot wallet to generate receiving addresses for both wallets, but it can only spend its own funds, i.e., it cannot spend the funds in the cold wallet. This design conveniently mimics money storage in daily life: pocket money is kept in a less safe location, e.g., a regular wallet, while life savings are kept in a more safe environment, e.g., banking account. Note that the funds that land in offline addresses cannot be spent if the cold wallet is kept permanently offline. We propose a protocol and a technical solution to spend funds from a cold wallet without physically connecting it to any network. We designed and implemented a prototype for a system based on Optical Camera Communication (OCC) in a screen to camera setting, which can receive messages from a computer screen at the rate of over 150kB per second. Our system consists of a sequence of QR codes – a QR video. Our solution minimizes the possible attack vectors, including malware, by relying on optical communication yet providing a larger bandwidth than regular QR code based solutions.