Cogensec: Frontier AI Security Lab

Cogensec

About Cogensec

Advancing the science of agentic security.

Cogensec is a frontier research lab building enterprise products for securing, governing, and advancing safe autonomous AI. We publish what we learn and ship what we prove.

Read our research Get in touch

research · products · standards THE LAB

Our Mission

Integrity is the only durable control.

We build products that protect autonomous AI systems at scale, securing agent identity, preserving evidence integrity, and enforcing governance across the entire AI lifecycle.

Perimeters were designed for software that waits to be told what to do. Agents don’t wait. Our research and our platforms start from the same premise: security has to be endogenous: carried inside the system that acts, not bolted around it.

Est. 2025

Founded

Product platforms

59k

Jailbreak corpus records

Remote

Research-first team

Founders

Built by practitioners, not observers.

Security researchers and engineers who spent careers defending real systems before turning to the ones that act on their own.

Mansoor Malik

CEO & Co-Founder

Network infrastructure and security architecture leader. Previously drove security initiatives across government and federal agencies, building programs where failure was not an acceptable outcome.

Tarique Smith

CTO & Co-Founder

Security researcher and product leader. Trained law enforcement agents in digital forensics and led product & application security at Verizon. Builds security platforms that hold up at enterprise scale.

Areas of Focus

Securing the full lifecycle of autonomy.

How Cogensec secures agentic and autonomous AI systems across the full lifecycle of autonomy.

Protect who acts

Agent Identity & Access

Protect who and what is allowed to act within your AI systems through robust authentication and permission controls.

Operating Principles

What we hold constant.

Four principles that decide what we build, what we publish, and what we refuse to ship.

01: Judgement

Human-Centered AI

Autonomy without accountability is a liability. Our systems augment the operators who carry the risk, keeping human judgement in the decision loop.

02: Frontier

Continuous Innovation

The attack surface moves weekly. We invest ahead of the threat curve, new attack classes, new detection primitives, new governance mechanics.

03: Field-led

Customer Obsession

Every capability we build traces back to a real incident, a real audit finding, or a real operator constraint from the field.

04: Privacy

Privacy by Design

Data sovereignty is a default, not a configuration. Local-first execution, minimal retention, and compliance-grade controls from the first commit.

Engagement Models

How Organizations Work With Us

Choose the engagement model that fits your security maturity and operational needs.

AI Security Readiness Assessments

Time-boxed evaluations of exposure across identity, data, tools, and decision logic.

Best for

Organizations beginning AI adoptionRegulated industriesPre-deployment validation

Adversarial AI Red Teaming

Live attack simulations against production or staging environments.

Best for

Production AI systemsHigh-risk automationCompliance requirements

Continuous Protection Platform

Integrated security controls embedded directly into AI pipelines.

Best for

Enterprise-scale deploymentsMission-critical AIOngoing security monitoring

Government & Critical Infrastructure

Mission-aligned deployments focused on resilience, integrity, and national security.

Best for

Public sector organizationsDefense contractorsCritical infrastructure operators

coverage sweep

Why Cogensec

Building Trust Through Research Excellence

Our platform is built by a team of AI security specialists, compliance officers, and governance experts with decades of combined experience securing enterprise systems. We understand the unique challenges of AI security because we’ve lived them.

Research Excellence

Frontier security research

Publishing peer-reviewed research on AI security, threat detection, and governance frameworks, and taking that work into industry standards.

Enterprise Products

Research shipped as product

Translating research breakthroughs into production-ready security products that protect organizations at scale. From agent discovery to real-time guardrails.

Open Innovation

Open tools, frameworks and datasets

Contributing to open-source security tools, frameworks, and datasets to elevate the entire industry’s security posture. Building together, securing together.

Thought Leadership

Teaching the field

Educating the industry through conferences, workshops, and security academies on emerging AI threats and defenses. Knowledge shared is security multiplied.

research in motion

Careers

Join the research lab.

A research-first organization where you have the autonomy to explore novel approaches to AI security while shipping products enterprises depend on.

Why work with us

Small team, unusually high leverage. The work you publish becomes standard; the systems you build run in production at organizations that cannot afford to be wrong.