Certiv
A Seattle startup that installs a runtime assurance layer on endpoints to discover and police what AI coding agents are allowed to do.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
Certiv builds Certiv Scout, a lightweight endpoint agent deployed through existing MDM tooling that sits between AI coding agents (Claude Code, GitHub Copilot, OpenAI Codex, Cursor and similar tools) and the operating system. It discovers which agents and local models are running on a device, including unsanctioned “shadow” agents, and evaluates each tool call, shell command, file access or API request against policy before it executes, rather than authorizing a session once and trusting it thereafter.
The company was founded in Seattle in mid-2025 by CEO Jason Needham (previously co-founder of Union Bay Networks, acquired by Apple, and CloudCoreo, acquired by VMware), CTO Paul Allen and Chief AI Officer Daniel Morris. It emerged from stealth in March 2026 with $4.2 million in pre-seed funding from Aviso Ventures, Founders Co-op and Fortson.
Its differentiator is working at the OS/endpoint layer across Windows, Mac, Linux and containers rather than inside any single coding tool, giving security teams one control point for policy enforcement, continuous per-action authorization, and audit trails mapped to frameworks like OWASP and MITRE ATT&CK.
Innovation Matrix Assessment
Went from founding (June 2025) to a funded, multi-platform product with reported pilot deployments within months of emerging from stealth in March 2026.
Directly addresses a live CISO problem: enforcing policy on what Claude Code, Copilot and similar coding agents can actually do on an endpoint, not just whether they're allowed to run.
Only $4.2M in pre-seed funding and nine employees as of launch; pilot deployments are mentioned but no named enterprise customers were found.
A genuinely new control point (pre-execution, per-action policy enforcement at the OS layer across agent tools) in a fast-forming category, but it competes with several other new entrants targeting the same AI-agent-runtime-security niche.
No independent testing or named-incident evidence was found in this research; claims of pilot success are self-reported by the company.
AI coding agents are rapidly becoming permanent parts of software development workflows, making endpoint-level governance of their actions a durable need.
Why CISOs Should Care
Gives security teams a single policy-enforcement point for AI coding agents across devices, instead of relying on each agent vendor's own (often limited) guardrails.
What Makes It Different
Enforces policy at the OS/endpoint layer, pre-execution and per-action, across multiple agent tools at once, rather than inside any single agent's own sandbox.
The Matrix Verdict
60/100 — INCREMENTAL INNOVATOR
An Incremental Innovator: a credible, fast-moving answer to a real and growing problem, but still pre-seed scale with no independent efficacy evidence yet.
Editorial Note: Claims vs. Verified Findings
Pilot deployment claims and policy-enforcement effectiveness are company-reported; no independent test or named customer outcome was found.
Sources
Alternatives to Certiv
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
CISO ReviewedEarly-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Gray Swan AI
Carnegie Mellon-linked startup building automated red-teaming and runtime guardrails that frontier AI labs use to stress-test their models…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…