Claude Code

AI is not fundamentally changing adversary capabilities but is compressing attack timelines, lowering operational costs, and scaling existing tactics. Breakout times have dropped to an average of 29 minutes, with AI-enabled operations increasing 89% year-on-year. The most significant emerging threats are runtime-LLM malware (PROMPTSTEAL/LAMEHUG, QUIETVAULT) that query language models during execution, and agentic AI operations (GTG-1002) where AI agents conduct multi-stage intrusions with minimal human steering. Defenders face a dual pressure: faster attacks and an expanding attack surface from AI supply-chain dependencies.

2026-06-036 minTrail of Bitshigh

Security researchers successfully bypassed multiple AI agent skill scanners, including ClawHub, Cisco's skill-scanner, and skills.sh integrations, using techniques like file truncation, embedded payloads, Python bytecode poisoning, and prompt injection. The findings highlight that automated scanning of AI agent skills is fundamentally flawed due to the complex mix of natural language, code, and limited scanner context windows, necessitating a shift towards curated, trusted skill repositories.

2026-05-265 minCheck Pointhigh

During March-April 2026, threat actors increasingly deployed commercial AI models for real-time offensive operations, including automated intelligence analysis, BEC drafting, and vulnerability exploitation. Key developments include the weaponization of agentic configuration files for persistent jailbreaks, the rise of AI-integrated PhaaS platforms like EvilTokens, and the mass harvesting of AI provider credentials. Furthermore, AI capabilities are compressing the vulnerability patch window, allowing attackers to weaponize newly disclosed CVEs within hours.

2026-05-156 minAkamaicritical

The TeamPCP threat actor deployed the Mini Shai-Hulud worm in a sophisticated supply chain attack targeting the npm ecosystem via a GitHub Actions CI cache-poisoning technique. The malware steals credentials, establishes persistence via developer tools like VS Code and Claude Code, and features a destructive dead man switch that wipes the victim's home directory if access tokens are revoked.

2026-05-135 minHuntresscritical

A Huntress engineer encountered a malvertising campaign via a Google sponsored search result for 'Claude Code'. The malicious link delivered a multi-stage macOS malware utilizing base64 encoding, gzip compression, and obfuscated AppleScript to bypass Gatekeeper and attempt extraction of Claude Code credentials from the macOS keychain.

2026-05-134 minElastic Security Labsinfo

Elastic's InfoSec team details a scalable architecture for monitoring AI coding assistants, specifically Claude Code and Cowork, using OpenTelemetry and Elasticsearch. The solution provides security teams with critical visibility into AI agent activities, including shell command execution, file access, and internal API interactions, enabling advanced threat detection, incident response, and EDR correlation.

2026-05-133 minTrail of Bitsinfo

Trail of Bits has released Trailmark, an open-source library that converts source code into queryable call graphs to enhance AI-assisted security analysis. By integrating with Claude Code, Trailmark enables advanced mutation testing triage, blast radius analysis, and the identification of architectural bottlenecks in cryptographic libraries.

2026-05-135 minTrend Microhigh

Threat actors are actively exploiting the recent Claude Code packaging error by hosting fake GitHub repositories that distribute trojanized archives. These archives contain a Rust-compiled dropper that deploys Vidar, GhostSocks, and PureLog Stealer to harvest credentials and establish residential proxies on compromised Windows systems.

2026-05-137 minTrend Microhigh

Following an accidental source code leak of Anthropic's Claude Code via npm, threat actors rapidly deployed fake GitHub repositories to distribute a Rust-compiled dropper. This dropper, part of a broader rotating-lure campaign, deploys Vidar stealer and GhostSocks proxy while utilizing extensive anti-analysis checks and PowerShell to disable Windows Defender.

2026-05-134 minVaronishigh

The source code for Anthropic's Claude Code CLI was accidentally exposed through .map files in a public npm release. This leak reveals the internal architecture, permission models, and safety guardrails of the AI agent, potentially allowing attackers to craft targeted prompt injections or distribute tampered dependencies through unofficial repositories.

2026-05-135 minTrend Microhigh

Following an accidental leak of Anthropic's Claude Code source material, threat actors rapidly deployed a social engineering campaign using fake GitHub repositories. The campaign distributes trojanized archives containing a Rust-compiled dropper that deploys Vidar stealer and GhostSocks proxy malware, specifically targeting developers seeking AI tools.

2026-05-134 minZscaler ThreatLabzhigh

Anthropic accidentally leaked the source code for its Claude Code CLI tool via an npm package source map. Threat actors are exploiting the high interest in this leak by creating fake GitHub repositories that distribute a Rust-based dropper, which subsequently installs Vidar infostealer and GhostSocks proxy malware on developer workstations.

2026-05-133 minTrail of Bitslow

Trail of Bits details their organizational shift to an AI-native workflow using Claude Code and autonomous agents. The post outlines their strategy for overcoming employee resistance, establishing an AI Maturity Matrix, and securing agent autonomy through sandboxing, curated marketplaces, and strict usage policies.

2026-05-133 minElastic Security Labslow

Elastic has introduced open-source Agent Skills that enable AI coding agents to natively interact with Elastic Security. These skills allow security teams to rapidly provision cloud environments, generate realistic sample attack data, and manage alerts and detection rules directly from their IDEs.

2026-05-134 minTrend Microhigh

TrendAI researchers demonstrated novel attack vectors against AI systems, including exploiting AI-driven KYC pipelines using 'executable documents' to leak customer data. Additionally, they introduced FENRIR, an automated vulnerability hunting system that has discovered numerous zero-days in AI and Model Context Protocol (MCP) ecosystems.

2026-05-135 minCheck Pointcritical

Check Point Research discovered critical vulnerabilities in Anthropic's Claude Code CLI that enable Remote Code Execution (RCE) and API token exfiltration. By injecting malicious configurations into project files like .claude/settings.json and .mcp.json, attackers could execute arbitrary commands and steal API keys when a developer opens a compromised repository, leading to potential supply chain attacks and unauthorized access to shared Claude Workspaces.