Activity timeline
AML.T0051 first appeared in tracked threats on and was most recently reported on . The busiest month was 2026-10 with 9 reports, and 33 of the 33 threats were reported in the twelve months to 2026-11.
How adversaries use it
AML.T0051 LLM Prompt Injection is catalogued by MITRE ATLAS under the Execution (ATLAS) tactic in the ATLAS matrix. Threadlinqs maps 33 of 3980 tracked threats (0.8%) to it; by severity that is 6 critical, 14 high, 13 medium.
Threats that use AML.T0051 most often also use T1027 Obfuscated Files or Information (17 threats), AML.T0054 LLM Jailbreak (11 threats), T1566 Phishing (11 threats), T1204 User Execution (10 threats), T1567 Exfiltration Over Web Service (9 threats). These are the techniques an intrusion set tends to chain with it, so they are the natural next places to look when it is observed.
2 tracked threat actors appear in the threats that use AML.T0051; the most frequent are Cleaver (2), TeamPCP (1).
Threat actors using it
Tracked threats
The 30 most recent of 33 tracked threats that use AML.T0051.
- Software Supply Chain Attacks Weaponizing Trusted Developer Workflows (Shai-Hulud, Miasma, Hades, TeamPCP)high
- OpenClaw apply_patch Path Traversal via LLM Guardrail Bypass (CVE-2026-32060 / GHSA-r5fq-947m-xm57)high
- Claude Dispatch "ThisBash" flaw lets malicious browser extensions drive Claude Desktop to execute commands…high
- Haiou Jailbreak Bundle: 'Prompt Adware' Persistently Altering AI Coding Agent Instructions to Promote LLM…medium
- Amazon Kiro IDE Prompt Injection Enabling Data Exfiltration via Malicious Workspace (CVE-2026-89332)medium
- W32/SkyAI (Topozuy/Skynet) Windows malware embeds AI prompt injection to evade AI-assisted malware analysismedium
- OpenAI API 'Compaction Smuggling' Flaw Exposes Hidden GPT-5.6 Reasoning Tracesmedium
- NVIDIA NemoClaw Chat Template Poisoning via Exposed Ollama API and DNS Rebinding (CVE-2026-65105)high
- Autonomous LLMjacking: 'Skynet' Agent Hijacks Exposed AI Models to Run Attackshigh
- Malicious OpenClaw AI Agent Skills: Reverse Shells, Semantic Worm, SSH Key Injection, Credential…high
- DHS Bans DeepSeek AI Models and Products from Federal Civilian Agencies (Foreign-Adversary AI Supply-Chain…medium
- ChatGPT GitHub Connector Indirect Prompt Injection Allows Private Repository Secret Exfiltration via…critical
- Attackers Hide AI Prompt Injections Inside Phishing Emails to Manipulate AI Email Assistantsmedium
- Coordinated model-distillation campaign against OpenAI: 15,000+ accounts attempt to extract protected model…medium
- Phishing Sites Engineered to Deceive AI Agents via Hidden Machine-Readable Instructions (Indirect Prompt…medium
- AI-Powered Cyber Attacks: Emerging TTPs Across Phishing, Deepfake BEC, Polymorphic Malware, and Prompt…medium
- SalesBleed: Salesforce Agentforce vulnerabilities enable zero-click CRM data theft and trusted-agent Slack…high
- Attackers Manipulate AI Chatbots (ChatGPT, Gemini, Copilot, AI Overviews) via SEO/Content Poisoning for Mass…high
- China-Based AI Companies Conducting Industrial-Scale Distillation Campaigns Against U.S. Frontier AI Modelscritical
- China-Based AI Companies Conducting Industrial-Scale Knowledge Distillation Campaigns Against U.S. Frontier…high
- Autonomous AI-agent frameworks automating credential theft and cyber espionage (Google Threat Intelligence…high
- Slopsquatting: Attackers Weaponize AI-Hallucinated Package Names in Supply Chain Attacksmedium
- Qilin-Linked Campaign Exploits MCP Gateway and LLM Framework Flaws (CVE-2026-59822, CVE-2026-42271…critical
- Critical Type Confusion in isolated-vm (GHSA-864f-rcv7-6rh4) Enables Sandbox Escape and RCE on Hostcritical
- Critical Microsoft Copilot CoSnitch Vulnerability (CVE-2026-24301) Enabled One-Click Data Theft From…critical
- Operation ASTERIX: AI-Assisted Crypto Wallet Phishing/Vishing Fraud Pipeline Abuses Claude Code and Kimihigh
- CSS Bomb Attacks: CSS-Based Trust-Boundary Bypass Leaks Webmail Passwords and Tokens (Outlook, Gmail, Yahoo…high
- CSS Bomb: JavaScript-Free CSS Keylogging and Token-Theft Attacks Against Gmail, Outlook, Yahoo Mail, AOL…high
- RovoBlast: One-Click Parameter-to-Prompt Injection in Atlassian Rovo AI Exposes Confluence, Jira, and…critical
- OWASP GenAI LLM Top 10 2026 — Community-Driven Security Guidance for AI Applicationsmedium
Detection coverage
Threadlinqs maintains 64 detection rules mapped to AML.T0051 (SPL 20, KQL 18, Sigma 26). Rule content unlocks with TL Creds on a free account; this page shows counts only.
Sub-techniques
- AML.T0051.000 LLM Prompt Injection: Direct — 2 tracked threats
- AML.T0051.001 LLM Prompt Injection: Indirect — 19 tracked threats
- AML.T0051.002 Triggered — 1 tracked threat