Topic

Anthropic: news, advisories and analysis

10 articlesUpdated 1 Nov 2026

CyberSecureToday covers Anthropic across 10 articles published between 17 Sep 2026 and 1 Nov 2026. The coverage includes 6 vulnerability advisories, 1 incident or breach report and 3 analysis or guidance pieces. 6 advisories are rated critical or high severity.

Stories on Anthropic most often intersect with AI Security, Claude and Model Context Protocol, with Technology, Cloud & SaaS the most affected sector. The most recent story, “Critical Linux Kernel IPv4 TCP Use-After-Free Flaw Enables Remote Root RCE (CVE-2026-98432)”, was published 1 Nov 2026. Every article is listed below, newest first, with the affected products, severity and remediation guidance in each piece.

Latest coverage

ResearchCloud & SaaS
Critical Linux Kernel IPv4 TCP Use-After-Free Flaw Enables Remote Root RCE (CVE-2026-98432)
CVE-2026-98432, a critical 9.8 CVSS flaw in the Linux IPv4 TCP stack, permits unauthenticated remote code execution on servers with MPTCP listeners.
Mei-Lin Chen1 Nov 202614 min read
ThreatsCloud & SaaSCritical
Trojanized Claude Code Extension on Cursor Marketplace Injects Reverse SSH Tunnel Backdoor
Attackers published backdoored versions 2.1.312 and 2.1.313 of the Claude Code extension on the Cursor marketplace to deploy persistent reverse SSH tunnels.
Mei-Lin Chen31 Oct 202611 min read
ResearchCloud & SaaS
Claude Solves 25-Year NIST Elliptic Curve Mystery, Debunking NSA Backdoor Theories
Anthropic's Claude reconstructed the 1997 Unix timestamps behind NIST ECDSA curve seeds, proving standard curves were randomly generated without an NSA backdoor.
Priya Raman18 Oct 202612 min read
ThreatsCloud & SaaSCritical
Claude Agent SDK Argument Injection Flaw Enables Command Execution via MCP Configuration
Anthropic resolves a critical argument injection defect in Claude Agent SDK (CVE-2026-96620, CVSS 9.2) that allows untrusted session parameters to execute arbitrary OS commands.
Aisha Noor17 Oct 20268 min read
ThreatsCloud & SaaSCritical
Anthropic Model Context Protocol Server Flaw Enables Host Directory Traversal and Unauthenticated Tool Injection (CVE-2026-70114)
Anthropic issues advisory CVE-2026-70114 for the Model Context Protocol (MCP) filesystem server, patching a path traversal flaw that allowed tool-calling escapes.
CST Editorial Desk11 Oct 20268 min read
ThreatsCloud & SaaSCritical
Anthropic Model Context Protocol Flaw Permits Indirect Prompt Injection Across Claude Enterprise Tool Callers
Anthropic and open-source MCP maintainers resolve CVE-2026-49110, an indirect prompt injection vulnerability allowing unauthorized tool execution in agentic pipelines.
Aisha Noor8 Oct 20268 min read
Cyber AttackGov & Defence
China-Aligned TA419 Impersonates White House and Anthropic Figures to Target AI Experts
Proofpoint: China-aligned TA419 posed as an ex-White House OSTP official and an Anthropic employee to steal AI policy experts' Microsoft 365 sessions.
Daniel Okafor3 Oct 20266 min read
ThreatsCloud & SaaSCritical
Hacktron Uses Claude Opus 5 to Beat libheif ASLR, Infiltrate OpenAI Perimeter
Hacktron AI researchers demonstrate how Claude Opus 5 synthesized an ASLR-bypassing libheif exploit to breach OpenAI forum and pivot to internal repos.
CST Editorial Desk21 Sep 202614 min read
ResearchCloud & SaaS
Agentic AI Security Breakdown: How Indirect Prompt Injections Hijack Autonomous Tool-Calling Agents
Analysis of emerging frontier AI threat vectors: how indirect prompt injections in untrusted documents hijack tool-calling autonomous LLM agents.
Priya Raman17 Sep 202614 min read
ThreatsCloud & SaaSCritical
Anthropic Threat Report: State-Sponsored Actors Exploit Claude for Automated Malware Evasion
Anthropic threat intelligence reveals nation-state groups, including Midnight Blizzard, attempted to weaponize Claude to automate malware obfuscation and EDR evasion.
Mei-Lin Chen17 Sep 202611 min read