AI Briefing
KOSign in

Cogent deploys Cyber Defense Factories using OpenAI models to cut vulnerability remediation time by 75%

·2026.11.10 09:00

Key point

The startup combines GPT-6.1 Sol and GPT-5.6 with enterprise context to autonomously investigate and fix security risks, achieving 5x triage throughput for IT services clients.

1 / 4

Details

  • Cogent Security builds Cyber Defense Factories that use OpenAI models to autonomously find, investigate, and remediate enterprise vulnerabilities, addressing the complexity of fixing security flaws across multiple teams.
  • A Fortune 1000 automotive customer reduced mean time to remediate critical vulnerabilities by 75%, while a Fortune 500 IT services customer achieved 5x vulnerability-triage throughput.
  • The system utilizes a Context Engine to maintain an enterprise knowledge graph and Agent Policies with a deterministic enforcement engine called TRACE to check tool calls against permitted changes before execution.
  • Cogent employs GPT-6.1 Sol for general execution and complex reasoning, and GPT-5.6 for routine vulnerability management, citing higher fidelity and lower costs compared to other providers.
  • The company uses Codex for internal development and accesses specialized models through Daybreak Red for authorized offensive research and attack-path analysis.
openaicodexcybersecurityai-agentsvulnerability-managementgpt-5.6cogentgpt-6.1-sol
Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Related

OpenAI BlogAmplitude's Wave agent uses OpenAI models to ship product improvements in 1–2 daysOpenAI BlogCult.fit standardizes on Codex to cut support costs 50% and save $100,000 annuallyRedditUnit 42 Reports Chinese Attackers Used DeepSeek and GLM Agents to Breach 10 OrganizationsRedditOpenAI internal cyber model hits 77.0% on ExploitGym, nearly double Anthropic's Mythos 5.2GeekNewsAnthropic and OpenAI Replace Git Systems Due to AI Agent LoadTLDR AIUnited Airlines to restrict AI agent access over security and regulatory risks