| AI Agent Benchmarks are Broken | Agent Security | Agent Security Frameworks and Tools |
| Benchmarks to compare Copilot adoption coming to Copilot Dashboard | Coding Tools | Coding Tools and Assistants |
| Microsoft adds Copilot adoption benchmarks to Viva Insights | Coding Tools | Coding Tools and Assistants |
| Indirect Prompt Injections: Are Firewalls All You Need, or Stronger Benchmarks? | Papers | Prompt Injection and Jailbreaking |
| Inside the Benchmark: App Architectures, Finding Walkthroughs, and What Each Scanner Actually Caught | Benchmarks | Security and Risk Benchmarks |
| N-Day-Bench | Benchmarks | Security and Risk Benchmarks |
| AI Bug-Bounty Agent Benchmark: 10 Models, 100 Black-Box Labs | Benchmarks | Security and Risk Benchmarks |
| Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study | Benchmarks | Security and Risk Benchmarks |
| Benchmarking 13 AI Models on Known CVE Detection | Benchmarks | Security and Risk Benchmarks |
| bots-bench - Benchmarking AI Models and Agents for SOC and IR Investigations | Benchmarks | Security and Risk Benchmarks |
| BoxPwnr - benchmarking LLMs and agents on security CTF challenges | Benchmarks | Security and Risk Benchmarks |
| Cisco LLM Security Leaderboard | Benchmarks | Security and Risk Benchmarks |
| Cotool BlueBench Windows Enterprise Intrusion benchmark | Benchmarks | Security and Risk Benchmarks |
| CTI-REALM: A new benchmark for end-to-end detection rule generation with AI agents | Benchmarks | Security and Risk Benchmarks |
| CVE-Bench - Benchmarking LLMs on real-world CVE patching | Benchmarks | Security and Risk Benchmarks |
| ExploitGym: Can AI Agents Turn Security Vulnerabilities into Real Attacks | Benchmarks | Security and Risk Benchmarks |
| FusionBench - Benchmark of AI Security Scan Models on a Fixed Vulnerability Corpus | Benchmarks | Security and Risk Benchmarks |
| Introducing dfbench v1: A cybersecurity benchmark for frontier models and agentic systems | Benchmarks | Security and Risk Benchmarks |
| Introducing EchoBench: A Human-Calibrated Benchmark for Autonomous Pentesting - NetSPI | Benchmarks | Security and Risk Benchmarks |
| JailbreakBench - LLM jailbreak robustness benchmark | Benchmarks | Security and Risk Benchmarks |
| LLMs Cannot Reliably Detect Vulnerabilities in JavaScript: The First Systematic Benchmark and Evaluation | Benchmarks | Security and Risk Benchmarks |
| PATCHEVAL: A New Benchmark for Evaluating LLMs on Patching Real-World Vulnerabilities | Benchmarks | Security and Risk Benchmarks |
| PWNBench: AI Pentesting Benchmark for 11 Frontier LLMs | Benchmarks | Security and Risk Benchmarks |
| Safety and Security Analysis of Large Language Models: Benchmarking Risk Profile and Harm Potential | Benchmarks | Security and Risk Benchmarks |
| SecureVibeBench: Benchmarking Secure Vibe Coding of AI Agents via Reconstructing Vulnerability-Introducing Scenarios | Benchmarks | Security and Risk Benchmarks |
| SnitchBench - AI whistleblower benchmark visualizer | Benchmarks | Security and Risk Benchmarks |
| Snyk VulnBench - Benchmark of how reliably AI systems find vulnerabilities | Benchmarks | Security and Risk Benchmarks |
| Sol Searching - Can Frontier Models Tackle Autonomous Long-Horizon Malware Analysis | Benchmarks | Security and Risk Benchmarks |
| T2I-RiskyPrompt: A Benchmark for Safety Evaluation, Attack, and Defense on Text-to-Image Models | Benchmarks | Security and Risk Benchmarks |
| TeleAI-Safety: A comprehensive LLM jailbreaking benchmark towards attacks, defenses, and evaluations | Benchmarks | Security and Risk Benchmarks |
| We burned 11.7bn tokens to find the best cyber AI model - GLM5.3 and DeepSeek are now frontier | Benchmarks | Security and Risk Benchmarks |
| We have Mythos at Home: GLM 5.2 beats Claude in our Cyber Benchmarks - Semgrep | Benchmarks | Security and Risk Benchmarks |
| BattleBench | Benchmarks | Evaluation Frameworks |
| Microsoft Foundry risk and safety evaluations (preview) Transparenc | Benchmarks | Evaluation Frameworks |
| Minimum Viable Benchmark | Benchmarks | Evaluation Frameworks |
| Benchmarking GPT-5.1 vs Gemini 3.0 v Opus 4.5 across 3 Coding Tasks | Benchmarks | Evaluation Frameworks |
| Alpha Arena - AI Trading Benchmark | Benchmarks | Evaluation Frameworks |
| LLMs are beating regex in secret detection\. We benchmarked Gitleaks, TruffleHog, and | Benchmarks | Evaluation Frameworks |
| LLM Leaderboard - Comparison of over 100 AI models from OpenAI, Goo | Benchmarks | Evaluation Frameworks |
| Long-form Content: AI (AI is Anti-Human (and assorted qualification | Benchmarks | Evaluation Frameworks |
| Scheming reasoning evaluations — Apollo Research | Benchmarks | Evaluation Frameworks |
| eyeballvul: a future-proof benchmark for vulnerability detection in | Benchmarks | Evaluation Frameworks |
| Did You Train on My Dataset? Towards Public Dataset Protection with | Benchmarks | Evaluation Frameworks |
| Arize Phoenix - open-source AI observability and evaluation | Benchmarks | Evaluation Frameworks |
| Dave Kennedy on Model Regression - daily benchmark testing of Claude, GPT, and Grok | Benchmarks | Evaluation Frameworks |
| Evaluating Large Language Models' Abilities to Process and Understand Technical Policy Reports | Benchmarks | Evaluation Frameworks |
| MatrAIx: Simulating the World with 8.3 Billion Persona Agents | Benchmarks | Evaluation Frameworks |
| ModelRegression.com - AI model performance and regression tracker | Benchmarks | Evaluation Frameworks |
| Optimal stopping: spending evaluation compute where it counts | Benchmarks | Evaluation Frameworks |
| Run evaluations from the Microsoft Foundry portal | Benchmarks | Evaluation Frameworks |
| SAGE: A Generic Framework for LLM Safety Evaluation | Benchmarks | Evaluation Frameworks |
| The Human Creativity Benchmark | Benchmarks | Evaluation Frameworks |
| DoomArena: A framework for Testing AI Agents Against Evolving Security Threats | Benchmarks | Agent Evaluation |
| CyberGym: Evaluating AI Agents' Cybersecurity Capabilities with Real-World Vulnerabilities at Scale | Benchmarks | Agent Evaluation |
| Benchmarking Practices in LLM-driven Offensive Security: Testbeds, Metrics, and Experiment Design | Benchmarks | Agent Evaluation |
| CAIBench - A Meta-Benchmark for Evaluating Cybersecurity AI Agents | Benchmarks | Agent Evaluation |
| CryptoAnalystBench: Failures in Multi-Tool Long-Form LLM Analysis | Benchmarks | Agent Evaluation |
| Enclave: We Raced Seven AI Models to RCE | Benchmarks | Agent Evaluation |
| Evaluating Skills, Not Just Agents: Agentic Continuous Evaluation of Skills | Benchmarks | Agent Evaluation |
| Mine the Gap: Open-Source Tools for Measuring the AI Offense-Defense Gap | Benchmarks | Agent Evaluation |
| SafeAgentBench: A Benchmark for Safe Task Planning of Embodied LLM Agents | Benchmarks | Agent Evaluation |
| SWE-CI: Evaluating Agent Capabilities in Maintaining Codebases via Continuous Integration | Benchmarks | Agent Evaluation |
| Grok 4.6 returns SpaceXAI to the intelligence frontier and leads on cost efficiency | General Reading | AI Industry and Ecosystem |
| iruletheworldmo's GPT-6 rumor thread claiming April 14 release and AGI-level benchmarks | General Reading | AI Industry and Ecosystem |
| Kaggle - AI competitions, benchmarks and datasets platform | General Reading | AI Industry and Ecosystem |