Security Articles
10 articles across AI Catchup's news, guides, tutorials, and comparisons.
All Security articles
Claude Managed Agents Add Memory Stores, Domain Controls, and a Redesigned Console
Anthropic's Claude Managed Agents updates let self-hosted sandbox sessions attach memory stores, restrict web_search and web_fetch with allowed_domains or blocked_domains, and inspect multi-agent sessions in a redesigned Claude Console viewer.
OpenAI Expands Daybreak With GPT-5.6-Cyber
OpenAI expanded its Daybreak cybersecurity program with Blue and Red access tiers and introduced GPT-5.6-Cyber, a purpose-trained model for authorized vulnerability research, exploit validation, and security testing. Access is limited to approved defenders and organizations.
OpenAI Brings Codex Security Review to GitHub Pull Requests
Codex Security Review is OpenAI’s research-preview workflow for deeper security analysis of GitHub pull requests. It uses the diff, repository context, and optional threat-model guidance, then reports actionable findings in the pull request and a fuller report in Codex. Enterprise, Business, Edu, and Pro users can configure it; Plus is excluded.
OpenAI Releases Codex Security CLI for Repository Scans and CI Checks
OpenAI says its open-source Codex Security CLI can scan repositories, track findings across runs, verify fixes, and add security checks to CI/CD. The beta CLI requires Codex Security access and is built for teams that want code-aware security review in the terminal.
Google Ships Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
Google released three Gemini models on July 21, 2026: Gemini 3.6 Flash, a coding and knowledge-work workhorse at $1.50/$7.50 per million tokens that uses 17% fewer output tokens; Gemini 3.5 Flash-Lite, a 350 tokens-per-second model at $0.30/$2.50 for high-throughput agents; and Gemini 3.5 Flash Cyber, a security-specialized model piloted with governments and trusted partners.
Codex Permission Profiles: Least-Privilege Controls for Local Agent Work
OpenAI shipped Codex permission profiles in beta -- reusable, inheritable policies that replace the coarse sandbox_mode/sandbox_workspace_write combo. A profile binds OS-enforced filesystem read/write/deny rules (down to **/*.env) to per-domain network and Unix-socket rules. Enterprise admins get fail-closed allowlists via requirements.toml. Profiles govern local sandboxed command execution only, not MCP servers, app connectors, browser, or cloud.
OpenAI's Secure MCP Tunnel: Connect Private MCP Servers Over Outbound-Only HTTPS
OpenAI's Secure MCP Tunnel connects private and on-prem MCP servers to ChatGPT, Codex, and the Responses API without opening inbound firewall ports. You need a tunnel_id, an API key with Tunnels Read plus Use, and -- separately -- ChatGPT developer-mode access. That permission split is the most common setup blocker.
Claude Code ships a security-guidance plugin for in-session vulnerability checks
Anthropic shipped an official security-guidance plugin for Claude Code. It runs automatic vulnerability checks while Claude edits files, at the end of each turn, and when Claude runs commits or pushes through its Bash tool.
OpenAI Codex adds Locked computer use on Mac (keep Computer Use running after lock)
Codex can now keep using Mac apps after your screen locks. Locked computer use is a narrow, Codex-only unlock path with a short-lived authorization window and safeguards like relocking on local input.
Claude Managed Agents Add Self-Hosted Sandboxes (Public Beta) and MCP Tunnels (Research Preview)
Anthropic says Claude Managed Agents can now run tool execution in a sandbox you control (public beta) and connect to private MCP servers via MCP tunnels (research preview). The update targets enterprise security requirements by keeping execution and private services within an organization's perimeter.