Machine learning techniques receive significant responsibilities, despite growing privacy concerns. Early-stage autonomous vehicles are increasingly appearing on the streets, carrying the burden of transporting human-lives to their destination. Meanwhile, doctors are involving Artificial Intelligence (AI) in their medical diagnoses, basing treatment of patients on the analyses AI provides. For these services to reach their full potential, a vast amount of training data is required, often gathered from a variety of sources. In many cases, the required data is considered to be privacy-sensitive (e.g., medical data). Due to the sensitivity of the underlying information, many individuals and organizations are not willing to entrust its protection to another party.
A field that attempts to limit the need to transfer training data openly is called collaborative learning, where multiple data generators cooperate to jointly train a classifier. In the proposed techniques the participants aim to limit the privacy loss of their collected training data to other collaborators. We contribute a clear overview of the current state-of-the-art and identify its limitations. Based on these limitations, we present two innovative protocol designs that pave the way towards private collaborative learning.
The ECONoMy protocol is developed to suit the needs of a high participant use case (i.e., Internet of Things (IoT)), under an assumed semi-honest adversarial model. The experimental results show that ECONoMy offers the desired privacy properties while remaining competitive to the non-privacy preserving alternative with which it is compared. However, in certain environments the incentives can grow exceedingly large rendering the ’semi-honest’ adversary assumption impractical.
We, therefore, created the PRECLUDE protocol which uses traceable ring signatures to protect against adversaries in the covert adversarial model. The tracing capability allows to detect malpractice and leak the identity of the deviant while preserving the anonymity of honest participants. These additional privacy-preservation properties came at a high cost to the overall efficiency, which is what we aimed to reduce by designing our extended protocol called PRECLUDE+.
PRECLUDE+ manages to drastically improve efficiency by reducing the number of participants included in a single signature. Further, we created a batch-verification phase that allows us to omit several exponentiations in each execution. We provide a detailed statistical analysis showing how to balance the efficiency improvements, with the required privacy parameters. The protocols presented in this thesis significantly improve upon the privacy guarantees offered by current alternatives, and provide a clear direction in which future work can continue to build.

Within the phenomenon known as the Internet of Things (IoT), an enormous growth is taking place. IoT systems exist in different ways, ranging from industrial applications to user focused systems. A specific subset of a user-focused IoT system is found as Smart Home environments. At Smart Homes, themultiple Smart Objects or Smart Devices are working together, frequently based on sensor input, to increase the comfort and user experience of the home inhabitant(s) and guest(s). Smart Objects can have automated tasks, home security enabling functions or efficiency improving functionality. Apart from great applications of Smart Home devices, threats from a cyber security perspective are present: cyber risks arise due to a variety of threats on such IoT systems. We show that in the development of new Smart Home products or systems, vendors fail to meet requirements for security and privacy are not met. Comparing the current state of the market, the four most used Smart Home ecosystems (Samsung Smartthings, Apple Homekit, Amazon Echo and IFTTT) are surveyed based on three key focus areas: 1. The regulatory compliance of the systems according to the upcoming General Data Protection Regulation (GDPR). 2. The commercial threats due to data profiling. 3. The risk of data leaks due to insufficient security. This analysis results in four key observations: 1. Security- and Privacy-By-Design is usually not in place due to the fact that the focus lies on launching a product as soon as possible, e.g. due to market competition; 2. Vendors process (meta)data on the vendors locations resulting in data profiling, which can compromise user privacy; 3. Smart Home ecosystems are not ready for the GDPR; 4. A trade off between privacy, security and utility usually results to the detriment of the first two and favors the latter. We propose a new design for a Smart Home ecosystem. In this design, the focus lies at the privacy of the end-user. We design a network for device-fitting encrypted communication between Smart Devices and User Devices and the Privacy Enforcing Arbiter (or Peter). Peter functions like a hub in the network, managing among others all traffic, user privileges and key distribution. With Peter, the centralized cloud party (vendor) for data storage and data analysis is replaced with a decentralized personal storage and computation entity at home. With our network design, we facilitate the use of IoT devices in home in a privacy-friendly way. Within the network, devices are authenticated using PhysicallyUnclonable Function technology and users are authenticated with a Zero Knowledge Proof. We analyze the privacy and security of our proposed network, based on a series of possible cyber attacks and the upcoming GDPR. Furthermore, we analyze the computational complexity and scalability of the network, based on market conform device power.