Even the largest tech companies suffer from data breaches, highlighting the need for a new approach to security like confidential computing.
A TEE is a secure, hardware-isolated enclave within a CPU that protects data while it is being processed in memory.
Confidential computing relies on data confidentiality, data integrity, and remote attestation to provide verifiable security for workloads.
Attestation is a cryptographic process that proves a TEE has not been tampered with before any sensitive data is sent to it.
Major hardware vendors like Intel, AMD, and Nvidia, along with cloud providers, offer a growing ecosystem of TEE-enabled products and services.
TEEs can be used to process PII, secure credentials, run private ML models, protect IP, and enable multi-party collaboration on sensitive data.
This demo shows how to run an ML model for brain tumor detection inside an AWS Nitro Enclave to protect both the model and patient data.
A simple mortgage application demonstrates how to isolate and protect sensitive financial calculations within a TEE while integrating with a standard web app.
While challenges like performance, cost, and side-channel risks exist, the goal is for confidential computing to become a ubiquitous security standard.
The speaker answers audience questions about successful implementations, integrating confidential computing into a security strategy, and the risks of AI in cybersecurity.
29:34Reto Kaeser
56:06Dwayne McDaniel
37:36Vandana Verma
40:38Anna Bacher
45:19Reinhard Kugler
45:19Philippe De Ryck
51:41Panel Discussion
37:32Martin Schmiedecker
Jobs that call for the skills explored in this talk.
SYSKRON GmbH
Regensburg, Germany
NTT DATA Deutschland GmbH
NTT DATA Deutschland GmbH