OpenAI Codex: two flaws allowed sandbox escapes
Two flaws fixed in OpenAI Codex allowed its sandbox to be bypassed through writes outside the workspace or a privileged component.
Read the article
Processing…
News
isolation, control and defense in depth to operate agents and generated code without opening infrastructure.
5 articles found
Two flaws fixed in OpenAI Codex allowed its sandbox to be bypassed through writes outside the workspace or a privileged component.
Read the article
OpenAI launches GPT-6 Astra, an AI capable of coding, browsing, using a computer, and carrying out complex tasks from start to finish.
Read the article
Google publishes Mantis, an open source framework that coordinates several AI agents to analyze code, verify vulnerabilities and propose fixes.
Read the article
Agent Identity becomes available to centralize identifiers, authenticate agents and better control their access to external tools.
Read the article
Google Cloud and Anyscale integrate gVisor to Ray sandboxes to isolate dynamic execution of post-training agents and workflows.
Read the articleEnter at least 2 characters to start searching.