GTGene Tsudik

Gene Tsudik

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Gene Tsudik has a total of 94 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Supporting Private and Efficient Consumer Requests to Data Brokers

    by Elina van Kempen, Pavel Frolikov, Chloe Georgiou, Gene Tsudik

    Under data privacy regulations, such as GDPR and CCPA, consumers are legally entitled to access, delete, or opt-out from collection of, personal information that so-called data brokers collect about them. However, the process of exercising these rights remains non-uniform, inefficient, and privacy-invasive. A consumer must submit individual requests to hundreds of data brokers, each requiring different - and often extensive - identity verification procedures. In January 2026, California launched a centralized facility that allows consumers to formulate a single (unified) request to all registered data brokers. While this simplifies the process, it raises significant privacy, scalability, and security concerns. Current commercial services (e.g., DeleteMe and Incogni) automate such requests for a fee, although they require consumers to provide all personal information upfront to yet another entity of uncertain trustworthiness. Motivated by the above, this paper presents PIVOT, an efficient cryptographic scheme that enables privacy-agile data access, deletion, and opt-out requests through an untrusted intermediary. PIVOT combines proxy re-encryption with additively homomorphic encryption so that data brokers can verify consumer identity without the intermediary learning any personally identifiable information (PII). A consumer encrypts their PII once, and the intermediary facilitates consumer identity verification with all data brokers. Data brokers learn only whether their databases contain matching records corresponding to a given consumer, without acquiring any new PII beyond what they already possess. PIVOT ensures both consumer and broker confidentiality, as well as verification soundness, requiring no user interaction after the initial request.

  • Open Access

    AuthentiSafe: Lightweight and Future-Proof Device-to-Device Authentication for IoT

    by Lukas Petzi, Torsten Krauß, Alexandra Dmitrienko, Gene Tsudik


  • Open Access

    Robust Secure Aggregation For Co-located IoT Devices With Corruption Localization

    by Giovanni Di Crescenzo, Elina van Kempen, Gene Tsudik

    Many Internet-of-Things (IoT) applications benefit from secure Federated (Machine) Learning (FL) techniques, e.g., industrial automation and smart cities. Since IoT devices have limited resources, such techniques need to be very efficient. In this paper, we propose a formal model as well as a new means for provably secure and efficient (i.e., IoT-friendly) aggregation for FL in IoT settings. Besides input privacy, it offers result correctness and malicious input localization, both robust against up to a threshold of malicious devices. The proposed techniques involve no interaction among devices, which only send a short message, and perform lightweight encryption and integrity detection.

  • Open Access

    TOCTOU Resilient Attestation for IoT Networks

    by Pavel Frolikov, Youngil Kim, Renascence Tarafder Prapty, Gene Tsudik

  • Open Access

    Acoustic Side-Channel Attacks on a Computer Mouse

    by Mauro Conti, Marin Duroyon, Gabriele Orazi, Gene Tsudik

    Acoustic Side-Channel Attacks (ASCAs) extract sensitive information by using audio emitted from a computing devices and their peripherals. Attacks targeting keyboards are popular and have been explored in the literature. However, similar attacks targeting other human interface peripherals, such as computer mice, are under-explored. To this end, this paper considers security leakage via acoustic signals emanating from normal mouse usage. We first confirm feasibility of such attacks by showing a proof-of-concept attack that classifies four mouse movements with 97% accuracy in a controlled environment. We then evolve the attack towards discerning twelve unique mouse movements using a smartphone to record the experiment. Using Machine Learning (ML) techniques, the model is trained on an experiment with six participants to be generalizable and discern among twelve movements with 94% accuracy. In addition, we experiment with an attack that detects a user action of closing a full-screen window on a laptop. Achieving an accuracy of 91%, this experiment highlights exploiting audio leakage from computer mouse movements in a realistic scenario.

  • Open Access

    MADEA: A Malware Detection Architecture for IoT blending Network Monitoring and Device Attestation

    by Renascence Tarafder Prapty, Rahmadi Trimananda, Sashidhar Jakkamsetti, Gene Tsudik, Athina Markopoulou

    Internet-of-Things (IoT) devices are vulnerable to malware and require new mitigation techniques due to their limited resources. To that end, previous research has used periodic Remote Attestation (RA) or Traffic Analysis (TA) to detect malware in IoT devices. However, RA is expensive, and TA only raises suspicion without confirming malware presence. To solve this, we design MADEA, the first system that blends RA and TA to offer a comprehensive approach to malware detection for the IoT ecosystem. TA builds profiles of expected packet traces during benign operations of each device and then uses them to detect malware from network traffic in real-time. RA confirms the presence or absence of malware on the device. MADEA achieves 100% true positive rate. It also outperforms other approaches with 160x faster detection time. Finally, without MADEA, effective periodic RA can consume at least ~14x the amount of energy that a device needs in one hour.

  • Robust Secure Aggregation for Co-located IoT Devices with Corruption Localization

    by Giovanni Di Crescenzo, Elina van Kempen, Gene Tsudik

  • Open Access

    Staving off the IoT Armageddon

    by Gene Tsudik

    IoT devices are increasingly popular and ubiquitous in numerous everyday settings. These specialized gadgets sense and actuate the environment using a wide range of analog peripherals. They are usually deployed in large numbers and often perform safety-and/or mission-critical tasks, in both military and civilian domains. It is no surprise that they represent attractive targets for various attacks. Adversaries range from nation-states to groups (motivated by politics, competition, or greed), to individual malcontents. Their goals vary based on targeted functionality: compromised sensors can exfiltrate sensitive information, while compromised actuators can affect the environment, i.e., physical safety and security. The well-known Stuxnet (2010) is an example of the latter, while numerous hacks into IoT cams exemplify the former. The infamous Mirai botnet (2017) is yet another "preview of coming attractions": it successfully zombified a huge number of consumer-grade cameras to form a global botnet later used to mount massive Distributed Denialof- Service (DDoS) attacks.1 Sadly, recent history shows that few, if any, lessons were learned as a result of these attacks. Although not quite malware-relevant, the recent CrowdStrike fiasco underscores the problem. IoT devices are still commonly compromised via both known attack types and zero-day exploits. Realistically speaking, the worst is yet to come. Unfortunately, the current security research limelight is on (both real and imagined) dangers of AI and Machine Learning algorithms, their unfairness, etc. There is thus a real risk of missing the real and present danger posed by the rampant (in)security of the IoT ecosystem. What makes the situation so dire? There are several reasons: First, most IoT-focused attacks and exploits are not physical in nature, meaning that they do not require the adversary to be present at or near victim devices. Because devices are increasingly interconnected and/or connected to the global Internet, they can be reached and attacked remotely. This yields many benefits to the adversary, such as much greater scale of attacks and relative impunity. Remote attacks occur because, similar to general-purpose computers, most IoT devices are programmable and/or configurable, meaning that they contain software/firmware that is fundamentally malleable in order to support legitimate updates and customizable functionality. Second, device manufacturers face certain constraints driven by the desire to offer new or better device functionalities at lower prices, and to do so in a hurry, i.e., the "rush-to-market" syndrome. Since security, in and of itself, is not viewed as a useful service, but rather as an inhibitor, it is either ignored or treated as an afterthought. This occurs due to lack of security expertise or because of the pressure to keep device costs low. It thus seems acceptable for manufacturers to improve security reactively, e.g., release patched software/firmware only after an attacks occurs or an exploit becomes public. Furthermore, security is viewed as a barrier to usability, making devices harder to set up and operate by users who are not technologicall adept.


  • Acoustic Side-Channel Attacks on a Computer Mouse

    by Mauro Conti, Marin Duroyon, Gabriele Orazi, Gene Tsudik