Google MCP Toolbox CVE-2026-9739 — DNS Rebinding to Enterprise Databases
CVSS 9.4 flaw in Google MCP Toolbox for Databases lets unauthenticated attackers abuse DNS rebinding via a hardcoded wildcard CORS header on the SSE transport.
High-signal AI/security/automation notes.
CVSS 9.4 flaw in Google MCP Toolbox for Databases lets unauthenticated attackers abuse DNS rebinding via a hardcoded wildcard CORS header on the SSE transport.
Mercor confirmed a cascading supply-chain attack through compromised LiteLLM packages, resulting in 4 TB of internal data exfiltration including Slack archives, source code, and contractor PII.
Microsoft identifies 14 malicious npm packages from actor vpmdhaj harvesting AWS, Vault, and GitHub Actions secrets from AI and developer environments.
OpenAI requires hardware-backed passkeys for Trusted Access for Cyber program users starting June 1, 2026, setting an identity-security precedent for AI operator access.
STAR Labs developed four complete exploit chains targeting LiteLLM across versions 1.82.3 to 1.83.14, escalating from internal user keys to server RCE.
Multiple CVEs across vLLM, MCP reference servers, and agent tooling expose AI inference and orchestration layers to RCE, DoS, and path traversal.
CISA adds CVE-2026-0257 to its KEV catalog after Rapid7 confirms active exploitation of the PAN-OS GlobalProtect authentication bypass.
The FBI warns of Kali365, a PhaaS platform using AI-generated lures and voice cloning to steal Microsoft 365 OAuth tokens via device-code phishing.
OWASP classified ASI06 memory poisoning in early 2026 — malicious instructions written to long-term agent memory that fire weeks later with the credibility of memory itself.
OWASP releases Agent Memory Guard, a drop-in middleware framework that wraps LangChain, LlamaIndex, and CrewAI memory APIs to defend against the ASI06 memory poisoning threat class.
Pentest Swarm AI integrates nmap, sqlmap, Burp, and Metasploit into an autonomous pentesting tool exposed as an MCP server.
Two new PraisonAI CVEs enable workspace takeover via missing member-removal authorization and cross-workspace label manipulation through unvalidated IDOR parameters.
Two vLLM CVEs: a multimodal heap-address leak weakening ASLR (CVE-2026-22778) and an unbounded n-parameter DoS in the OpenAI-compatible API (CVE-2026-34756).
Shanghai Jiao Tong University researchers introduce IterInject, an indirect prompt injection framework that uses LLM-driven iterative optimization to adapt payloads against agent defenses, achieving full success on 5 of 9 Claude Code targets.
Detectify launches an MCP Server that exposes its scanning engines to AI coding agents, enabling autonomous vulnerability finding, patching, and validation in real time.