Tagged articles
vulnerability detection
23 articles · Page 1 of 1
Google DeepMind released three new Gemini models—3.6 Flash with 65% token reduction, 3.5 Flash-Lite outperforming larger models at lower cost, and 3.5 Flash Cyber for vulnerability detection—while confirming the start of 'most aggressive' pre-training for Gemini 4, signaling cheaper, faster AI agents for production.
AI agentsGeminiGoogle DeepMind
0 likes · 8 min read

Cloudflare open-sources a six-phase security audit skill that transforms coding agents into structured auditors using adversarial validation, machine-readable findings, and a coverage ledger to make AI-driven vulnerability detection reproducible and trustworthy.
AI code reviewCloudflareCoding Agent
0 likes · 13 min read

The article details a 24‑hour AI‑driven red‑team/blue‑team workflow that continuously discovers hidden CDN bugs, generates protocol‑aware exploits, automatically repairs them, and continuously refines its knowledge base, achieving a 20‑fold speedup and over 99% fix rate.
AI-driven securityAgentic WorkflowCDN
0 likes · 34 min read

This article introduces the PHP Taint extension, explains its runtime taint‑tracking mechanism, shows how it marks user inputs, propagates taint through string operations, triggers warnings on high‑risk functions, and provides installation, usage examples, supported risk scenarios, built‑in APIs, and best‑practice recommendations for secure PHP development.
PHPSQL injectionTaint
0 likes · 10 min read

Anthropic's Claude Security, built on Claude Opus 4.7, entered open beta on April 30, offering enterprise AI‑driven code vulnerability scanning and automated patch generation, and completing a three‑product security suite that also includes Claude Code and Claude Cowork, with broader implications for blue‑team operations and AI‑enabled threat landscapes.
AI securityAnthropicClaude Security
0 likes · 8 min read

Released on July 8, 2026, xAI's Grok 4.5 outperforms competitors in software‑engineering benchmarks, consumes far fewer tokens, and has been praised by security researchers for its powerful vulnerability‑detection capability, while offering a $2/M‑input‑token pricing and 80 TPS inference speed.
AI ModelGrok 4.5benchmark
0 likes · 4 min read

A previously unknown Chinese AI security group called MopMonk, operating without a website or corporate backing, posted a GitHub report that achieved a 73.1% vulnerability‑exploitation success rate, ranked seventh globally in the UC Berkeley‑run CyberGym benchmark, and demonstrated novel memory‑based multi‑agent techniques that signal China’s rising AI security prowess.
AI securityCyberGymMiniMax M3
0 likes · 9 min read

The DCRH project is Anthropic’s production‑grade, open‑source reference implementation that leverages Claude’s large‑model multi‑agent architecture to build an end‑to‑end AI‑driven security pipeline, reducing false positives and speeding up vulnerability remediation for C/C++ codebases.
AI securityClaudeautomated remediation
0 likes · 9 min read

Analyzing 3,632 high‑severity GitHub Advisory reports from 2025‑2026, the authors reveal a sharp rise in business‑logic flaws—especially in high‑star projects—prompting a redesign of vulnerability‑detection benchmarks, and introduce VulnGym, a real‑project, white‑box dataset with 400+ paths and detailed entry‑point, trace, and critical‑operation annotations.
AI securityBusiness Logic Bugsbenchmark
0 likes · 17 min read

In just two weeks, Anthropic’s Claude Opus 4.6 model identified 22 security flaws in the Firefox codebase, including 14 high‑severity issues, demonstrating that advanced AI can move from auxiliary analysis to core vulnerability hunting and potentially reshape the security industry’s fundamental dynamics.
AI securityClaude OpusDevSecOps
0 likes · 6 min read

Anthropic's new Claude Code Security tool uses an AI model that reads code like a human researcher, detecting complex logic‑flaw and permission‑control bugs missed by traditional pattern‑matching scanners, providing multi‑round verification, confidence scores, and AI‑generated patches while still requiring developer approval.
AI securityAnthropicClaude
0 likes · 6 min read

This article explores how Wukong's AI‑driven multi‑agent architecture dramatically improves code security auditing by addressing context loss, scheduling imbalances, and integrating a data‑flywheel that turns bad cases into continuous model improvements, illustrated by a real NVIDIA Megatron‑LM vulnerability fix.
AICode AuditingSecurity
0 likes · 14 min read

XiaYue, a powerful Burp Suite extension, automates vertical and horizontal privilege escalation detection by comparing responses across multiple permission levels, offering smart deduplication, advanced filtering, parameter replacement, visual data tables, persistent configuration, and performance optimizations, while the author also shares a heartfelt story about their child's school start.
Burp SuiteWeb Securityautomation
0 likes · 9 min read

The A.S.E (AI Code Generation Security Evaluation) framework provides a comprehensive, project‑level benchmark for assessing the safety, quality, and stability of AI‑generated code across multiple languages and vulnerability types, helping developers and researchers evaluate and improve large language model coding assistants.
AI code generationopen-source benchmarksecurity evaluation
0 likes · 7 min read

This article analyzes common logical vulnerabilities in freight‑related APIs, compares SAST, IAST and DAST approaches, proposes a detection architecture with traffic collection, token handling, fuzzy‑hash comparison and API de‑duplication, and discusses remaining challenges and future improvements.
API SecurityDASTIAST
0 likes · 14 min read

The article reviews the rapid rise of Web3, outlines the severe security challenges and economic losses caused by attacks on cross‑chain bridges, DeFi platforms, and smart contracts, and details Ant Group's Sky Dome Lab research on platform, contract, and client‑side defenses along with future AI‑driven initiatives.
Blockchain SecurityDeFiSmart Contract
0 likes · 9 min read


AntTechApr 20, 2023 · Information Security
This study investigates the capability of ChatGPT, a large language model, to detect common smart‑contract vulnerabilities, evaluates its performance against traditional tools across multiple datasets, and discusses its strengths, limitations, and future potential in blockchain security auditing.
AIBlockchain SecurityChatGPT
0 likes · 23 min read


AntTechApr 19, 2023 · Information Security
The paper presents ODDFuzz, a structure‑aware directed greybox fuzzing framework that combines lightweight static taint analysis with targeted fuzzing to efficiently discover previously unknown Java deserialization (ODD) vulnerabilities, achieving higher recall and precision than existing tools and uncovering six new CVE‑rated bugs in popular Java frameworks.
FuzzingJavaODDFuzz
0 likes · 19 min read

The 2022 Vivo Qianjing Cup Cybersecurity Challenge, launched on September 15 under the “Assemble! Geek Youth” theme, invites nationwide security enthusiasts to uncover vulnerabilities in Vivo products through online puzzles and an offline final, offering a 200,000‑yuan prize pool plus uncapped bounty rewards and judging by experts from AWS, Ant Group and leading universities.
cybersecurityinformation securityprivacy protection
0 likes · 5 min read

This guide introduces CodeQL, explains how to install the required tools, shows how to generate a source‑code database, and walks through basic and advanced rule syntax with practical C/C++ examples, enabling security researchers to efficiently discover vulnerabilities in large codebases.
C++CodeQLCodeQL Queries
0 likes · 13 min read

This article explains what OS command injection is, how it can be detected and exploited on both Linux and Windows systems, demonstrates common payloads and techniques—including blind and out‑of‑band methods—and provides best‑practice defenses to prevent such vulnerabilities.
OS command injectiondefense techniquesshell injection
0 likes · 10 min read

The article explains why and when to scan product code for vulnerabilities, describes static source‑code and binary scanning methods, introduces deep code‑search techniques, outlines the system architecture and incremental indexing pipeline, and shows how these practices can substantially raise overall product quality.
Code ScanningProduct QualitySphinx
0 likes · 13 min read

This article outlines the motivation, design principles, architecture, component choices, and step‑by‑step implementation of a comprehensive server security audit system, covering server information collection, log gathering, access control checks, local vulnerability detection, abnormal traffic analysis, and integration with ELK, Hadoop, and open‑source tools like Lynis and OSSEC.
LynisOSSECSecurity Audit
0 likes · 20 min read
