USENIX Security ’25 (Enigma Track) – Trusted Hardware For Al Workloads: Extending Confidential Computing To Enable Al Adoption

[…Keep reading]

Author, Creator & Presenter: Shannon Egan, Deep Science VenturesAs companies race to adopt AI in new use cases, hardware vendors and cloud providers are developing the protocols to secure AI workloads with limited input from the broader security community. This talk surveys key challenges of extending Confidential Computing and Trusted Execution Environments from CPUs to clusters of AI accelerators, highlighting technical contributions needed from security experts: efficient remote attestation and key management, secure interconnects, and device memory protection. These advancements would enable stronger security guarantees while maintaining performance and code compatibility–crucial requirements for commercial adoption. We draw from our experience evaluating market opportunities for emerging technologies to offer a unique perspective on both the commercial potential and technical feasibility of trusted hardware for large-scale AI.

Our thanks to USENIX Security ’25 (Enigma Track) (USENIX ’25 for publishing their Creators, Authors and Presenter’s tremendous USENIX Security ’25 (Enigma Track) content on the Organizations’ YouTube Channel.

Permalink

*** This is a Security Bloggers Network syndicated blog from Infosecurity.US authored by Marc Handelman. Read the original post at: https://www.youtube-nocookie.com/embed/MLPrgiR5VQM?si=2Ggw2hnP9FzFWGuo

About Author

AndyC

Andy Curtis is an award-winning security consultant, researcher and public speaker. He has been working in the computer security industry since the early 1990s, having been employed by state and federal government, leading healthcare and banking providers across three continents. He has given talks about computer security for some of the world’s largest companies, worked with law enforcement agencies on investigations into hacking groups, and is a regular voice on TV and radio explaining IT security threats.

See author's posts