🧰 GPT-5.5

The items that name it 33

• How we detect frontier model misalignment
  articleas classification.technologies2026-10-29
• Red-Teaming OpenAI's Internal Agent Monitoring Systems
  writeupas classification.technologies2026-10-22
• Putting models to the secure coding test: Security rules
  writeupas classification.technologies2026-10-19
• ORCAGen: Orchestrating Context-Aware Malware Deception with RAG-Guided Generative AI
  paperas tools2026-10-08It is written here as GPT-5.5 (LLM evaluated; fewest refinements, no observed hallucinated APIs).

  LLM evaluated; fewest refinements, no observed hallucinated APIs

• 3 lessons from frontier AI vulnerability research
  writeupas tools2026-10-07It is written here as GPT-5.5 (model used for automated validation evaluations without task-specific tuning).

  model used for automated validation evaluations without task-specific tuning

• GPT-6 Sol and GPT-6 Luna: October 2026 update
  articleas classification.technologies2026-10-07
• Top AI agent security resources — October 2026
  articleas affected products2026-10-04It is written here as GPT-5.5 (OpenAI).

  OpenAI

• New AISI Report Details How GPT-6 Astra Turned CTF Challenges Into Supply Chain Attacks
  articleas classification.technologies2026-09-30
• Addendum to GPT-6 Astra System Card: GPT-6.1 Sol
  articleas affected products2026-09-29
• VulContextBench: A Benchmark for Security Context Retrieval in Coding Agents
  paperas tools2026-09-26

  Frontier model evaluated on the benchmark; views 73% of gold context and reports 36%, the highest citation rate among models tested

• 1Password's AI patching benchmark is misleading
  articleas affected products2026-09-15
• What is an agent harness? The software that turns a model into an agent | Blog | Endor Labs
  articleas affected products2026-08-15
• GPT-5.6 — August Updates
  writeupas classification.technologies2026-08-06
• AI Security Briefing
  articleas classification.technologies2026-07-18
• AI Security Briefing
  articleas classification.technologies2026-07-18
• AI Security Briefing
  articleas affected products2026-07-18

  model tested; leaked .env when driven through vulnerable coding tools

• OpenAI Codex with GPT-5.6 Sol: competitive, zero cheating, one unique Django fix | Blog | Endor Labs
  articleas tools2026-07-17

  Prior model generation tested via Codex, used for improvement comparison (also referenced as Cursor + GPT-5.5 combo scoring 24.0% SecPass)

• Benchmarking 13 AI models on rediscovering known CVEs
  articleas tools2026-07-16

  OpenAI model benchmarked; one of only two models (with GPT-5.6 sol & terra) that traced the SQL Injection-1 indirect injection chain

• XBOW: The Rise of Affordable Models
  articleas tools2026-07-16

  Frontier AI model evaluated for cybersecurity capability, noted as excelling especially at white-box testing

• XBOW: Grok 4.5 Poised To Take Over the Middle of the AI Security Market
  articleas tools2026-07-16

  AI model compared for offensive security performance; eventually reaches slightly higher max solve rate at high spend

• Claude Sonnet 5 with Claude Code: strong on function, average on security, and unusually honest | Blog | Endor Labs
  writeupas affected products2026-07-09
• Benchmarking Endor Labs AI SAST: 2.6x more real vulnerabilities found than frontier models | Blog | Endor Labs
  articleas classification.technologies2026-07-06
• GPT-5.5-Cyber built a zlib fuzzing lab in a day
  writeupas classification.technologies2026-07-02
• Introducing Patch the Planet
  articleas classification.technologies2026-06-22
• Endor Labs: Claude Fable 5
  writeupas classification.technologies2026-06-18
• When a Government Pulls an AI Model: What the Fable 5 and Mythos 5 Suspension Means for Security Teams
  articleas affected products2026-06-14

  Cited by Anthropic as offering the same level of code-analysis/remediation capability that was the reported trigger for the Fable 5/Mythos 5 suspension

• Endor Labs: AI Coding Agents Cheat Security Benchmarks via Training Recall, Not Reasoning
  articleas affected products2026-06-11
• The sorry state of skill distribution
  writeupas affected products2026-06-03It is written here as GPT 5.5.

  Default guard model used by ClawHub's custom scanning harness

• Move over, Mythos. Here comes... pretty much any other model with a good harness
  articleas tools2026-06-01

  Frontier AI model; benchmarked by the UK's AI Security Institute at roughly the same cyber capability tier as Mythos, reaching 71.4% on the hardest category of their evaluation suite versus Mythos's 68.6%.

• JFrog: Keep your Agents Under Control with agent-belt
  toolas affected products2026-05-19
• GPT-5.5 Vulnerability Detection Eval
  articleas tools2026-05-13

  OpenAI model evaluated by the UK's AI Security Institute for its ability to find security vulnerabilities; noted as generally available

• XBOW: GPT-5.5 and Opus 4.7 in Offensive Security Workflows
  writeupas tools2026-05-06

  LLM tested by XBOW across offensive security benchmarks and workflows, including vulnerability detection, black box and white box testing, computer use/login tasks, and persist-or-pivot decision making

• Google, Anthropic, and OpenAI Unveil Cyber AI Models, Safeguards, and Access Programs
  articleas classification.technologies

In every workspace you can read 70