agent security
Stories from agent security come first, newest first, then stories from connected topics (prompt injection, AI Agents, sandboxing, AI Agents, MCP, AI security), each labelled with where it comes from. This view is not indexed by search engines.
ServiceNow Forms 11-Member AI Alliance With OpenAI and Anthropic
ServiceNow's Control Tower Alliance has 11 founding members, including OpenAI, Anthropic, NVIDIA and Okta, to certify AI tools for governance.
Latest
Microsoft Says AI Agents Are Insider Risks, Adds Guardian Agents
Microsoft says frontier models "cannot be trusted", though not because they are malicious, and adds Guardian Agents to Project Perception to watch agents.
Snyk Launches Evo Agentic AppSec, a Team of AI Security Agents
Snyk launched Evo Agentic AppSec. Most new agents are in preview; VulnBench and VulnBump figures are Snyk’s claims, not independent audits.
Noma Labs Details “SneakPeek”, a Now-Fixed Flaw That Let Claude Code Desktop’s Preview Pane Read Local Secrets
Noma Labs says a now-fixed Claude Code Desktop flaw let a poisoned file in the preview pane read local secrets. Anthropic rated it CVSS 8.7, High.
Singapore Launches a Starter Kit for Testing AI Agents and a Task Force on Frontier-AI Cyber Threats
IMDA has issued a voluntary starter kit for testing AI agents and set up FAIRS, a task force on cyber threats from frontier AI, per CNA and ST.
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.
Grok Agents, Part 2: Tools, Connectors and Your Own Data
Part two of our Grok agent guide: tools, connectors and MCP in plain words, a safe read-only first setup, how to limit access and a short safety checklist.
What Grok Bot Connects To and What It Can Do Inside Each Service
What Grok Bot connects to, how sign-in and custom MCP connections work, what it can do inside Gmail, Drive, Notion and more, and how to limit its access.
AI Agent Security, 8 October (Evening): Disputed AgentCore Research, Claude Code’s Fail-Closed Hooks and Malware Aimed at AI Analysts
Three agent-security stories from 8 October. Zenity says one prompt to an exposed AWS AgentCore agent could lead to control of every agent in the same account and region, and AWS disputes it. Claude Code adds an option that blocks an action when a safety hook fails. Cisco Talos finds malware that writes to AI analysts. What each shows, and what it does not.
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.
Latest on TSN
Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.
