About

Closed Loop Security Labs

AI security research and services, powered by sovereign infrastructure.

Built by a Security Engineer

Closed Loop Security Labs was founded by Brandon Arbour, a network security engineer transitioning into AI security with deep experience in enterprise infrastructure, zero-trust architectures, and adversarial testing methodologies.

Prior to founding CLS Labs, Brandon served on Verizon's Executive Shadow Board, mentoring C-suite executives on generative AI strategy. That experience revealed a critical gap: organizations deploying AI at scale have no standardized way to verify that their security controls actually work.

CLS Labs exists to close that gap — literally. The closed loop is the product: attack, verify, remediate, prove.

Brandon Arbour, Founder of CLS Security Labs

Certifications

  • CompTIA Security+
  • CompTIA Pentest+
  • CompTIA CASP+
  • CompTIA CySA+
  • Google ACE
  • CompTIA SecAI+ (2026)

The Sovereign AI Factory

CLS Labs operates a privately-owned AI research cluster for sovereign workloads — no third-party data exposure, no cloud dependencies for sensitive operations. For clients requiring maximum data control, assessments run entirely on-premises. For scale and speed, we also operate across trusted cloud infrastructure with strict data handling policies. Your risk profile determines the deployment model — not our convenience.

Compute: Multi-GPU NVIDIA cluster for parallel inference and training
Inference: Multi-framework serving with unified API gateway
Orchestration: Kubernetes with eBPF-based network policy enforcement
Identity: Zero-trust identity with workload attestation
Monitoring: Full observability stack with SIEM integration
Networking: Segmented network with deep packet observability
Security: Proprietary and open-source adversarial testing tooling
Policy: Triple-Lineage Rule (no shared DNA in validation)

Audit & Compliance Stack

Every assessment produces a complete chain of evidence from probe to remediation.

Data Store: Versioned, append-only data lake with transactional backing store
Classification: Proprietary Semantic Defense Engine
Logging: SIEM with immutable audit trails
Provenance: CLAP Layer 2 verification pipeline (N≥3 reproductions)
Frameworks: NIST AI RMF • OWASP LLM Top 10 • MITRE ATLAS
Guardrails: Multi-stage integrity verification pipeline
Config Mgmt: Signed, versioned, rollback-capable guardrail configs

Contact

Email: info@clsecuritylabs.com
Web: clsecuritylabs.com
GitHub: github.com/clsecuritylabs