PuzzleMask Hides Attacks Inside Plain Prose Text
A small model reads a text and answers “safe.” Then the very same text reaches a more powerful AI, which finds a hidden command within and executes it. Check Point Research has laid bare...
Tagged: LLM
The Illusion of Sapience: Unmasking the “Performative” AI and the Rise of Agentic Malware
Malefactors are already endeavoring to weave artificial intelligence into the fabric of malicious software, yet the current manifestations remain profoundly erratic. In certain instances, the neural network serves merely as an ostentatious facade, leaving...
The Infinite Variant: How “Promptmorphism” Uses AI to Shatter Traditional Malware Detection
Malefactors are increasingly harnessing large language models to rapidly rewrite malicious code. This stratagem, christened “promptmorphism,” facilitates the virtually infinite synthesis of novel initial-stage loaders. Such a tactical maneuver profoundly confounds the detection of...
The Hunter Becomes the Hunted: “Evil AI” WormGPT Suffers Massive Leak of 19,000 User Accounts
The WormGPT platform, notoriously recognized as an AI-augmented instrument for orchestrating cyber-offensives and infiltrations, has purportedly suffered a catastrophic data breach. On a prominent forum dedicated to the dissemination of compromised databases, an archive...
Robin: The AI-Powered “Sidekick” for Dark Web OSINT Investigations
Robin is an AI-powered tool for conducting dark web OSINT investigations. It leverages LLMs to refine queries, filter search results from dark web search engines, and provide an investigation summary. Features Modular Architecture – Clean...
Brute Force Reimagined: How BruteForceAI Uses LLMs to Crack Complex Login Forms
BruteForceAI is an advanced penetration testing tool that revolutionizes traditional brute-force attacks by integrating Large Language Models (LLM) for intelligent form analysis. The tool automatically identifies login form selectors using AI, then executes sophisticated...
AI Flunks Security Test: LLMs are Ineffective at Guessing Passwords from Personal Data
Australian researchers have tested whether large language models can infer passwords from personal information — and found that, for now, they are almost entirely ineffectual. In a new study, the team at the Future...
Apple Tightens AI Privacy: Apps Must Get Consent to Share Data with External LLMs
Apple has updated its App Store moderation rules, tightening oversight of how developers handle user data when working with third-party algorithms. The company has introduced a new requirement for apps that transmit information to...
SigmaOptimizer: End-to-End LLM Tool for Automated Sigma Rule Generation and Testing
SigmaOptimizer is a End-to-End Sigma rule generation and optimization tool that automatically creates, tests, and improves Sigma rules based on real-world logs using LLM. It is implemented as a PowerShell script and integrates log analysis, rule evaluation, and iterative...
Pentest Copilot: The New Open-Source AI Tool for Ethical Hacking
Pentest Copilot is an open-source tool built to assist ethical hackers and penetration testers. By integrating LLMs, it automates and enhances various pentesting tasks. The tool is deployable locally with Docker and includes an...
Next-Gen AI Security Engineers Arrive: New AI-SAST Tools Find Logic Flaws and 50+ Zero-Days in Code
The market has begun to see the emergence of the first truly functional solutions that harness artificial intelligence to detect vulnerabilities in source code. This new generation of AI-SAST systems — often referred to...
ASCII Smuggling Attack Bypasses Filters, Forcing Gemini to Obey Invisible Commands from Calendar Invites
Researchers at FireTail have discovered the resurrection of an old-class flaw — ASCII Smuggling — now resurfacing in a modern guise capable of subverting contemporary artificial-intelligence systems. In September 2025, the team evaluated a...
