Search
Samsung Developer Conference
Tech Session Advanced Tech, Security & Privacy, Open Sources, Platform

Islet: On-device Confidential Computing for Secure and Privacy-First AI

AI progress brings exciting opportunities for increased productivity and creativity, but involves vast amounts of personal data. Data misuse or leaks can result in severe privacy violations. Moreover, AI model parameters and algorithms are valuable intellectual property needing protection. Confidential Computing (CC) offers a promising approach to safeguard both AI models and user data on user devices. By applying CC, we can enhance security measures and ensure the safety of app codes and data in use. This opens up new possibilities for apps in various fields. In this presentation, I'll delve into how CC can be applied to user devices for AI model and user data protection. Furthermore, I'll explore the future potentials with on-device confidential computing and introduce Islet, an on-device CC platform developed by Samsung Research based on the new ARM CCA architecture. Islet is an official open source project under the Confidential Computing Consortium in Linux Foundation.

Speakers

Bokdeuk Jeong

Samsung Electronics

Back to List

Preferences Submitted

You have successfully updated your cookie preferences.