Awesome AI Agent Attacks

Slack AI exfiltration (Aug 2024): Markdown link rendering enables data exfiltration to attacker servers. ASCII smuggling M365 Copilot (Jul 2024): Invisible Unicode in hyperlinks carries

The New AI Attack Surface: 3 AI Security Predictions for 2026

Building your 2026 AI security roadmap requires confronting three attack vectors that are already manifesting in production environments.

Attack on Titan Revolution Codes [Update 4] (May 2026)

We''ve searched long and far across the internet to locate all of the currently working codes for Roblox Attack on Titan Revolution!

AI Cyberattacks: How attackers target AI, and use AI

Wiz''s research into agent frameworks such as MCP shows how over-privileged AI agents can be tricked into executing unintended tool actions,

Google

Checking your browser before accessing undefined Click here if you are not automatically redirected after 5 seconds. Checking your browser - reCAPTCHA

OWASP MCP Top 10

By mapping the top 10 MCP-related vulnerabilities and offering concrete recommendations for secure design, implementation, and auditing practices, this project aims to

Microsoft Exchange Zero-Day Under Attack, No Patch Available

Microsoft Exchange Zero-Day Under Attack, No Patch Available CVE-2026-42897 stems from a cross-site scripting (XSS) vulnerability and can allow an attacker to compromise Outlook Web

LLMjacking: what these attacks are, and how to protect AI servers

Consequently, when deploying proprietary AI servers and their supporting ecosystems like RAG or MCP, it''s critical to establish rigorous security measures from day one.

Attack on Titan Revolution (AOTR) Codes (May 2026)

Here is a list of new Attack on Titan Revolution codes that will grant you spins and luck boosts to become stronger and unlock better weapons.

Microsoft SQL Server Vulnerability Let Attackers

Attack Mechanism The vulnerability works by exploiting how SQL Server processes database names in queries. Attackers can craft malicious

The lethal trifecta for AI agents: private data, untrusted

The lethal trifecta for AI agents: private data, untrusted content, and external communication If you are a user of LLM systems that use tools (you can

Agentic AI Attack Surface: Why It''s the #1 Cyber Threat

48% of security professionals say agentic AI is the top attack vector for 2026. Learn why autonomous AI systems are expanding the enterprise attack

The Protocol Powering AI Agents Has a Hole

How the Attack Works MCP''s STDIO transport spawns a new process on the host machine every time an AI agent connects to a tool server. The protocol passes configuration parameters — including the

The Real-World Attacks Behind OWASP Agentic AI Top 10

OWASP''s new Agentic AI Top 10 highlights real-world attacks already targeting autonomous AI systems, from goal hijacking to malicious MCP

GitHub RCE Vulnerability: CVE-2026-3854 Breakdown | Wiz Blog

Securing GitHub: Wiz Research uncovers Remote Code Execution in GitHub.com and GitHub Enterprise Server (CVE-2026-3854) Details on CVE-2026-3854: A critical flaw in GitHub''s

Welcome to Channel Dive | Channel Dive

Welcome to Channel Dive. We''re Informa TechTarget''s new publication, focused on delivering daily news and analysis for executives at North

MCP Prompt Injection Attacks 2026: How Hackers Hijack AI Agents

AI agents using MCP are vulnerable to prompt injection, tool poisoning, and data exfiltration. Real attack examples + 7 defenses. MCP Security Gateway from $690.

MCP Server Security: The Hidden AI Attack Surface

Our research demonstrates that both locally hosted and third-party MCP servers can be exploited to execute arbitrary code, exfiltrate sensitive data,

MCP Prompt Injection Attacks 2026: How Hackers Hijack AI Agents

Quick Answer: MCP (Model Context Protocol) connects AI agents to external tools, databases, and APIs. In 2026, three attack vectors exploit this connection: (1) prompt injection

MCP Tool Poisoning: Enterprise AI Agent Security in 2026

MCP Tool Poisoning: Enterprise AI Agent Security in 2026 Tool poisoning has emerged as the highest-leverage attack on enterprise AI agents in 2026 — exploiting the metadata that agents

AI Kill Chain Simulation: LLM Exploit to Full Compromise

In this post, we walk through a fully simulated AI infrastructure compromise in which an attacker escalates from an initial foothold in an over-permissive LLM assistant to full control over a