Complete AI Training
Start for Free

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI news ·

Cogent Security Raises $42M to Automate Vulnerability Management with AI Agents, Bringing Total Funding to $53M

Cogent Security raised $42M led by Bain Capital Ventures, bringing total funding to $53M. Its agents automate vulnerability triage and remediation across existing workflows.

Cogent Security Raises $42M to Scale AI Agents for Vulnerability Remediation

Cogent Security announced a $42 million Series A led by Bain Capital Ventures, with participation from Greylock Partners, Definition, and executives from OpenAI, Abnormal Security, and Datadog. This brings the company's total funding to $53 million and will be used to accelerate product development.

The San Francisco-based company is building an agentic AI platform that automates the heavy lift in vulnerability management-investigation, prioritization, and remediation. It connects and normalizes data across environments, filters scanner noise, adds business context, and moves fixes through existing workflows. The goal is straightforward: shrink risk faster with less manual coordination.

"Security teams are drowning in coordination work: chasing down system owners, writing tickets, proving fixes happened. We built AI agents that handle that work end-to-end, so security teams can finally keep pace with attackers," said Vineet Edupuganti, CEO and co-founder of Cogent.

Why this matters for management

  • Turn vulnerability backlog into a predictable, measured pipeline of fixes.
  • Reduce coordination costs across security, IT, and dev teams without adding headcount.
  • Improve MTTR and SLA adherence by automating ticketing, ownership routing, and verification.
  • Prioritize what actually matters to the business, not just what has a high severity score.

How agentic AI can streamline your vulnerability operations

  • Consolidates and normalizes scanner output, config data, and asset inventories to reduce noise.
  • Prioritizes using business context and environmental signals, beyond CVSS alone.
  • Surfaces emerging threats and aligns them to affected assets and owners.
  • Generates remediation plans and pushes them into your existing ITSM/dev workflows.
  • Closes the loop by tracking proof-of-fix and exceptions with audit-ready evidence.

Questions to ask your team and the vendor

  • Data coverage: Which scanners, asset inventories, CMDBs, and ITSM tools are supported? How is data quality enforced?
  • Prioritization: What signals (business criticality, exploit activity, exposure) drive ranking? How are ties and conflicts handled?
  • Workflow control: What requires human approval vs. fully automated action? Can we set per-system guardrails and change windows?
  • Evidence: How is remediation verified (config drift checks, rescans, log proof)? Is evidence exportable for audits?
  • Security: How are credentials stored and scoped? Is there full audit logging, SSO, RBAC, and least-privilege access?
  • Reliability: What are the SLAs, rollout safety checks, and rollback paths if an automated change causes issues?
  • Deployment: SaaS vs. private deployment options? Data residency controls? Integration effort and typical time-to-value?

90-day pilot plan (practical and measurable)

  • Days 0-30: Connect one scanner, one CMDB, and one ITSM project. Define approval policy. Baseline metrics: backlog size, MTTR, ticket cycle time, false-positive rate.
  • Days 31-60: Enable automated ticket creation and owner routing for a narrow scope (e.g., one business unit). Track proof-of-fix capture and exception handling.
  • Days 61-90: Expand scope and introduce limited auto-remediation where safe. Compare results to baseline; decide on scale-up criteria and budget.

Metrics that prove value

  • Backlog reduction (%) and time to burn down criticals.
  • MTTR improvement (hours/days) by severity and asset class.
  • Percent of tickets auto-created, correctly routed, and auto-closed with verified evidence.
  • False-positive/duplicate reduction vs. prior quarter.
  • Analyst hours saved per week and avoided headcount growth.
  • SLA compliance rate and change-failure rate for remediations.

Risk and governance checklist

  • RBAC with approval thresholds; segregation of duties for creation vs. approval of fixes.
  • Comprehensive audit logs; evidence packages suitable for SOX/PCI/SOC 2.
  • Data handling: residency options, encryption, and no persistence of sensitive payloads in model training.
  • Guardrails: test in staging where possible; emergency stop and rollback paths; explicit change windows.
  • Hallucination and action limits: require validation steps for destructive or high-impact changes.
  • Third-party risk review and security assessment before broad rollout.

Budget and org impact

Expect savings from reduced coordination work, fewer reopens, and faster closure of high-risk items that usually stall in handoffs. Most teams reallocate analysts to higher-value engineering and threat work instead of adding headcount.

Model the ROI with conservative assumptions: ticket cycle time saved, percent of fixes verified without manual effort, and risk reduction for assets powering revenue or regulated workloads. Treat it as a platform investment-success depends on clean data, clear ownership, and well-defined approval rules.

Market signal

Backers like Bain Capital Ventures and Greylock leaning into agentic AI for vulnerability management signals growing confidence in automation beyond analysis-into action and verification. Useful for leaders tracking where security operations is heading and how to future-proof team structure and tooling.

Helpful references

For leaders setting AI-in-security strategy

If you're aligning AI initiatives with security governance, integration, and enterprise risk, see our AI Learning Path for CIOs for a structured approach to roadmapping, controls, and measurable outcomes.

Share