Nvidia has received rare US approval to export a small quantity of its H200 AI chips to China. While the move signals a potential shift in export policy, strict conditions, tariffs, and geopolitical tensions continue to limit real commercial impact.

AMD has proposed new Linux kernel patches to enable BTB isolation under SEV-SNP, adding another layer of protection for confidential virtual machines against branch prediction side-channel risks.

Meta is experimenting with stablecoin-based payments through third-party partners, marking a strategic shift in how the tech giant approaches digital currencies and in-app commerce.

Linus Torvalds has warned Linux developers against using unseeded randomness, calling it insecure and highlighting risks tied to weak entropy sources.

Qualcomm is pushing a new Linux DSP accelerator driver designed to standardize Hexagon DSP offload using the kernel’s accel framework.

The UAE has thwarted a coordinated wave of AI-enabled cyberattacks on critical infrastructure, underscoring the growing sophistication of modern digital threats.

Anthropic has introduced Claude Code Security, a new AI feature designed to uncover hidden vulnerabilities in software code and recommend targeted fixes.

A trusted VPN used across U.S. government networks failed to stop a sophisticated China-linked cyber intrusion, raising urgent questions about supply-chain security, third-party software risks, and the future of government cybersecurity defenses.