Fabricked: Exploiting Infinity Fabric Misconfigurations to Break AMD SEV-SNP Confidential Computing
By
negura
A good honest bake. Not flashy, but you'll finish the whole bagel.
Summary
This paper (Fabricked) presents a novel software-based attack that exploits misconfigurations in AMD's Infinity Fabric to break AMD SEV-SNP, a hardware-based trusted execution environment used for confidential computing. By redirecting memory transactions, a malicious hypervisor can deceive AMD's secure co-processor (PSP), compromising the security guarantees of Confidential Virtual Machines (CVMs) in cloud environments. The research was accepted at USENIX Security 2026.
Key quotes
· 3 pulledWith Fabricked, we present a novel software-based attack that manipulates memory routing to compromise AMD SEV-SNP.
By redirecting memory transactions, a malicious hypervisor can deceive the secure co-processor (PSP).
Confidential computing allows cloud tenants to offload sensitive computations and data to remote resources without needing to trust the cloud service provider.
You might also wanna read
Wi-Fi Router Beamforming Feature Can Be Exploited to Identify Individuals With 99.5% Accuracy, Study Finds
Researchers at Germany's Karlsruhe Institute of Technology discovered that standard Wi-Fi routers using beamforming feedback information (BF
futurism.com·5h ago
Behavioral feature engineering, not deep learning models, key to Trojan malware detection study finds
A study on Trojan malware detection focuses on behavioral feature engineering for Windows-based IoT and industrial systems. Rather than emph
helpnetsecurity.com·1d ago
MemoAttack: A Memory-Driven Framework for Automated LLM Jailbreak Attacks
This paper introduces MemoAttack, a novel memory-driven black-box jailbreak framework for large language models (LLMs). Unlike existing meth
arxiv.org·1d ago
CAPTCHAs remain viable for detecting AI agents by exploiting process differences
The article discusses how while AI vision language models (VLMs) can now solve traditional CAPTCHA image recognition tasks (like identifying
research.roundtable.ai·2d ago
Eureka: An LLM-Driven Framework for Automated Feature Engineering in Enterprise AI
This paper presents Eureka, an LLM-driven framework for automated feature engineering in machine learning. It treats feature engineering as
arxiv.org·4d ago
Researchers discover hidden audio signals can hijack AI voice systems
AI-powered voice and audio systems (large audio-language models) are increasingly used in daily life for voice commands, transcription, and
spectrum.ieee.org·13d ago