Check Point Research
PulseAugur coverage of Check Point Research — every cluster mentioning Check Point Research across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
MCP rug pulls: Servers alter tools post-approval, risking user trust
A security vulnerability known as an "MCP rug pull" allows malicious servers to alter their tool definitions after initial approval, potentially tricking users into executing harmful commands. This occurs because the MC…
-
AI to drive machine-speed cybersecurity race by 2027, experts predict
Cybersecurity is poised for a significant transformation in 2027, driven by the increasing autonomy of AI in both attack and defense strategies. AI-powered attacks will become more sophisticated and autonomous, necessit…
-
AI coding tools risk exposing API keys; new tool offers protection
AI coding assistants like Claude Code can inadvertently expose sensitive API keys by reading them from environment files and incorporating them into generated code or session history. This poses a significant security r…
-
LiteLLM Flaws Expose Admin Accounts and API Keys
Check Point Research has identified two critical vulnerabilities within the LiteLLM software. These flaws could allow authenticated users with low privileges to gain administrative control, access sensitive internal fil…
-
PuzzleMask AI Attack Hides Malicious Prompts from Safety Filters
Check Point Research has identified a new AI attack vector called PuzzleMask, which uses plain-prose prompts to conceal malicious instructions. This technique bypasses lightweight AI safety filters, with target models s…
-
AI Automation Tools Vulnerable to Sandbox Escapes and Data Theft
Critical vulnerabilities have been discovered in AI automation tools, including Cloudflare Code Mode, Google Gemini CLI, and Anthropic Claude Code. These flaws could allow attackers to escape sandboxes, steal data, and …
-
Exposed AI Endpoints Become New Attack Surface for Threat Actors
Threat actors are increasingly exploiting exposed AI endpoints, transforming them into offensive infrastructure. These endpoints, often integrated with sensitive internal systems like CRMs and code repositories, present…
-
AI Model DeepSeek Creates Novel Browser-Native Ransomware
A novel in-browser ransomware, dubbed "InfernoGrabber," has been independently created by the DeepSeek AI model. This ransomware operates entirely within the browser, utilizing the File System Access API to encrypt loca…
-
Critical vulnerabilities found in LangGraph AI framework
Researchers at Check Point Research have identified critical vulnerabilities within LangGraph, a framework for developing AI applications. These flaws, particularly in self-hosted deployments using SQLite and Redis, cou…
-
AI-driven cyberattacks are now routine, Check Point Research finds
Check Point Research's latest AI Threat Landscape digest reveals that AI-driven cyberattacks are now a common tool for criminals. The report highlights the use of malicious configuration files that bypass safety measure…
-
AI agents breach 600+ firewalls with autonomous cyberattacks
In early 2026, a series of nine coordinated cyberattacks, driven by LLM-powered agents, successfully breached over 600 enterprise firewalls. These autonomous systems discovered and exploited zero-day vulnerabilities at …
-
Vect's ransomware is a data wiper, making victim data unrecoverable
Cybersecurity researchers have discovered that the ransomware used by the Vect group, which has targeted numerous organizations since January, is actually a data wiper. This malware permanently destroys files larger tha…