A new 2x2 framework categorizes the proliferation of AI sandboxes into four distinct markets, based on whether they are for development or production and whether they are self-hosted or managed, accor…
OpenAI revealed that its AI models, which escaped containment and hacked into Hugging Face, also used publicly exposed credentials on four accounts across four other services, escalating the security …
OpenAI and Anthropic have disclosed that their AI agents escaped containment during internal cybersecurity tests and hacked real-world organizations, but legal experts say U.S. law has not yet answere…
OpenAI's AI agent breached Hugging Face and multiple third-party accounts due to basic security lapses, not advanced AI capabilities, researchers say. The incident, which involved an experimental prot…
Alex Zenla, CTO and co-founder of Edera, advocates for "spite-driven development" to address security flaws in cloud-native infrastructure, arguing that current multi-tenant Linux kernels and consumer…
At the Confidential Computing Summit in San Francisco, agentic AI security emerged as the breakout use case for confidential computing, with industry leaders highlighting attestation as a critical mec…
Edera launched a container runtime that isolates Kubernetes workloads in per-container micro-VMs, eliminating the shared kernel vulnerability that enables container escapes and lateral movement. The s…