Events featuring Carbyne Stack¶

Abstract

With speed, cost, and privacy advantages, federated learning is emerging as an alternative to centralized ML approaches. However, without additional safeguards, federated learning won't live up to its promise, as it is vulnerable to privacy attacks and model theft.

Our open-source Flower extension for privacy-preserving federated learning called Nettle fills this gap by deploying additional layers of defense. By integrating with our open-source Carbyne Stack Secure Multiparty Computation (MPC) platform, Nettle can effectively protect against privacy attacks via MPC-based Secure Aggregation. In addition, Nettle deploys Confidential Computing to protect valuable ML models from unauthorized model extraction on the client.

In this talk, we describe the concepts behind Nettle, which are based on the idea of splitting the role of the Flower server into the three distinct roles of model owner, orchestrator, and aggregator. We show how we leveraged Flower's extensibility mechanisms to enable straightforward integration with the Carbyne Stack MPC services providing a scalable backend for Secure Aggregation. We describe how Confidential Computing enclaves and remote attestation are used in Nettle to ensure that the model owner retains full control over the ML model throughout the whole distributed learning process. Finally, we give a preview on what is to come for Nettle and Carbyne Stack in the future.

Abstract

Computing on Encrypted Data (CoED) is considered a holy grail of data security. A major roadblock for the adoption of CoEDs is a lack of integration with cloud technologies to enable scalable, resilient, and easy to operate deployments. The Carbyne Stack open-source project has set out to close this gap. This talk will take the audience down the rabbit hole of CoED technologies and explain how Carbyne Stack blends cloud-native technologies to solve the challenges of scaling sensitive workloads.

Abstract

Data has become a strategic asset that is pooled with others for joint processing, monetized on data platforms, and used to fuel the AI revolution. As the ability to leverage internal and external data is becoming a major factor for business success, protecting valuable data is more important than ever. Enter Computing On Encrypted Data technologies (COEDs). COEDs pave the way for strong end-to-end protection of data by enabling encryption in use. One roadblock for the wider adoption of COEDs so far has been the lack of integration with state-of-the-art cloud technology to enable scalable, resilient, and easy to operate COED deployments. The Carbyne Stack open-source project has set out to close this gap by lifting a specific COED technology called Secure Multiparty Computation (MPC) into the cloud. Sven will take the audience down the rabbit hole of COED technologies and explain how Carbyne Stack blends cloud-native technology (including Kubernetes, Istio, Knative, and others) to solve the specific challenges of deploying MPC in the cloud like cross-cluster orchestration of MPC services and serverless provisioning of MPC workloads.

Abstract

Data has become a strategic asset that is pooled with others for joint processing, monetized on data platforms, and used to fuel the AI revolution. As the ability to leverage internal and external data is becoming a major factor for business success, protecting valuable data is more important than ever. Computing on Encrypted Data technologies in general and Secure Multiparty Computation (MPC) in particular pave the way for strong end-to-end protection of data by enabling encryption in use.

One roadblock for the wider adoption of MPC so far has been the lack of integration with state-of-the-art cloud technology to enable resilient, observable, and manageable MPC deployments at scale. The Carbyne Stack open-source project has set out to close this gap by lifting MPC into the cloud.

This talk will shed light on various aspects of Carbyne Stack: Why did Bosch Research start this initiative? How does the architecture and design of Carbyne Stack reflect the special characteristics of MPC? How can cloud-native technologies help in solving the challenges of scaling MPC workloads? In addition, the audience will get an outlook on what's next for Carbyne Stack and how interested individuals, institutions, and companies can contribute to the Carbyne Stack open-source project.

Abstract

Computing on Encrypted Data (CoED) technology is a new mind-boggling way of protecting data in use. CoEDs can be used to implement IT systems where sensitive data is encrypted end-to-end: in transit, at rest, and even in use. CoEDs will play an increasingly important role in the future when it comes to protecting sensitive information in collaborative environments or meeting the requirements of increasingly stringent, diverse, and pervasive data protection regulations in our target markets.

Participants of this tutorial will get to know the fundamentals of a CoED technology called Secure Multiparty Computation (MPC) that allows a set of mutually distrusting parties to securely compute a function of their private inputs, revealing only the output, even if some of the parties are corrupt. Participants will learn how to author simple MPC programs using a state-of-the-art MPC engine and how to use the Carbyne Stack open source cloud MPC platform to deploy MPC workloads in a scalable way. In addition, attendees will learn about the roadmap for the Carbyne Stack platform and how they can contribute to its development.