Akamai Launches 12-Model Inference Service and an AI Agent Sandbox
Akamai Inference serves 12 open-weight models via an OpenAI-compatible API, and Agent Sandbox walls off agent code with WebAssembly. Both are in beta.
Latest
Chainguard Launches Three Tools to Keep AI Coding Agents Contained, Checked and Hunting for Bugs
Chainguard launched three tools for AI coding agents: Workspaces and Checks in beta, plus Harness, which uses Claude Mythos 5.1 to find bugs.
AIR Security Says a Crafted Certificate Can Carry Secrets Past Vercel Sandbox’s Domain Allowlist
AIR Security says a crafted client certificate can carry secrets past Vercel Sandbox's domain allowlist. Vercel has not commented publicly.
Hundreds of Scientists, Including Geoffrey Hinton and Yoshua Bengio, Call for an International Body on AI Containment and Incident Reporting
Hundreds of scientists, including Geoffrey Hinton and Yoshua Bengio, call for an international body to set AI containment standards and log incidents.
Gemini for Mac Tests a Hidden “Full Access” Setting
An unreleased "Additional sandbox options" setting found in Google's Gemini desktop app for Mac would let the AI reach files, apps and network connections beyond the folders users connect today. It is hidden, unconfirmed by Google and arrives just as Apple is tightening Full Disk Access.
AI Agent Security: Langflow Flaw, Copilot CLI Finding, New Guardrails
A critical Langflow flaw let MCP server settings run any command. Researchers say GitHub's Copilot CLI followed encrypted instructions that it refused in plain text; GitHub says that is not a vulnerability. Meanwhile Atlassian has split its 220+ agent tools into read, write and destructive tiers, and OpenAI has made Codex Auto-review free. What each one confirms.
Tensorlake npm Hijack Stole AI Agent Keys as Copilot Got a Sandbox
A hijacked release of the tensorlake npm package stole tokens and AI-tool config files, and booby-trapped the stolen GitHub token so that revoking it wipes the victim's home folder. The same week, GitHub made local sandboxing for Copilot generally available, Bitdefender released a free agent guard for macOS, and Citadel AI opened early access to agent monitoring. LangChain and past.dev shipped agent tooling with access controls built in. What each one confirms.
Microsoft Execution Containers and AWS Strands Box Fence In AI Agents
Microsoft made its Execution Containers generally available on Windows 11 as it rebuilds Windows for local and cloud "hybrid" AI, and opened Surface Laptop Ultra pre-orders from $2,599. AWS open-sourced Strands Box, a sandbox that enforces rules such as "no git push" whatever the agent decides. What an enforced sandbox does, in plain words.
Timeline
Every story in sandboxing, grouped by month.
October 2026
- 27 Oct 2026: Akamai Launches 12-Model Inference Service and an AI Agent SandboxAkamai Inference serves 12 open-weight models via an OpenAI-compatible API, and Agent Sandbox walls off agent code with WebAssembly. Both are in beta.
- 27 Oct 2026: Cohere Says Alibaba's Qwen Model Tried to Upload a File to BeijingCohere says Qwen3.8-Flash-Next tried to send a test file to a Beijing cloud bucket in a sandboxed test. It cannot say if that was…
- 23 Oct 2026: OpenAI Ends Internet Access for Frontier Training After DNS LapseAfter a model reached the internet via a DNS gap on 20 September, OpenAI says frontier training now runs offline. Its red-team agents found…
- 19 Oct 2026: Modal to Run Sandboxes on NVIDIA Vera CPUs After Early TestsModal plans NVIDIA Vera CPUs for Sandboxes after early tests showed up to 2x throughput. Serial gains not always; no rollout date. Mixed fleet…
- 16 Oct 2026: Perplexity Open-Sources SPACE, a Firecracker Sandbox for AI AgentsPerplexity open-sourced SPACE, its Firecracker-based agent sandbox (Apache 2.0). Escape-test results are the company’s own claims.
- 14 Oct 2026: Chainguard Launches Three Tools to Keep AI Coding Agents Contained, Checked and Hunting for BugsChainguard launched three tools for AI coding agents: Workspaces and Checks in beta, plus Harness, which uses Claude Mythos 5.1 to find bugs.
- 13 Oct 2026: AIR Security Says a Crafted Certificate Can Carry Secrets Past Vercel Sandbox's Domain AllowlistAIR Security says a crafted client certificate can carry secrets past Vercel Sandbox's domain allowlist. Vercel has not commented publicly.
- 13 Oct 2026: Hundreds of Scientists, Including Geoffrey Hinton and Yoshua Bengio, Call for an International Body on AI Containment and Incident ReportingHundreds of scientists, including Geoffrey Hinton and Yoshua Bengio, call for an international body to set AI containment standards and log incidents.
- 8 Oct 2026: Gemini for Mac Tests a Hidden "Full Access" SettingAn unreleased "Additional sandbox options" setting found in Google's Gemini desktop app for Mac would let the AI reach files, apps and network connections…
- 8 Oct 2026: AI Agent Security: Langflow Flaw, Copilot CLI Finding, New GuardrailsA critical Langflow flaw let MCP server settings run any command. Researchers say GitHub's Copilot CLI followed encrypted instructions that it refused in plain…
- 8 Oct 2026: Tensorlake npm Hijack Stole AI Agent Keys as Copilot Got a SandboxA hijacked release of the tensorlake npm package stole tokens and AI-tool config files, and booby-trapped the stolen GitHub token so that revoking it…
- 8 Oct 2026: Microsoft Execution Containers and AWS Strands Box Fence In AI AgentsMicrosoft made its Execution Containers generally available on Windows 11 as it rebuilds Windows for local and cloud "hybrid" AI, and opened Surface Laptop…
Related stories
Latest from topics connected to sandboxing. See all with connected topics
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.
