# R6 Security — llms.txt

> R6 Security builds Phoenix, an Automated Moving Target Defense (AMTD) platform for Kubernetes and AI/ML workloads.

## About R6 Security

R6 Security is a cybersecurity company that developed Phoenix, the first infrastructure-native Automated Moving Target Defense platform. Instead of detecting attacks after they happen, Phoenix continuously rotates, mutates, and regenerates infrastructure so attackers never get a stable target. Founded by Zsolt Nemeth, the company is backed by an advisory board that includes former Gartner analysts, PhD researchers, and cybersecurity veterans.

## Phoenix Platform

Phoenix is a Kubernetes Operator that implements Automated Moving Target Defense through five mechanisms:

1. **Container Rotation** — Pods are replaced at random or telemetry-driven intervals, severing any attacker foothold.
2. **Runtime Mutation** — Environment variables, network identities, and service labels change continuously.
3. **Workload Regeneration** — Compromised workloads are automatically replaced with known-good baselines.
4. **Network Obfuscation** — IPs, service endpoints, and routing paths shift dynamically.
5. **Panic Mutation** — When Falco detects suspicious activity, Phoenix triggers immediate workload rotation.

### Key Facts

- Zero-agent architecture — no sidecars, no code changes, no instrumentation
- 1–2% infrastructure overhead (IBM Cost of a Data Breach Report 2024 estimates average breach cost at $4.88M; Ponemon Institute reports 277 average days to detect a breach — Phoenix makes both irrelevant by eliminating persistence)
- Deploys via Helm in under an hour
- Compatible with ArgoCD, Flux, Terraform, and standard GitOps pipelines
- Integrates with Falco for threat-signal-driven rotation

### Two Products

- **Phoenix for Kubernetes** — Pod rotation, dynamic label obfuscation, temporal network policies. Zero dwell time for Kubernetes clusters.
- **Phoenix for AI** — Inference pipeline mutation, NVIDIA NIM hardening, GPU resource protection. Prevents LLM-jacking and model theft.

## Published Research

- "ADA: Automated Moving Target Defense for AI Workloads via Ephemeral Infrastructure-Native Rotation in Kubernetes" — arXiv:2505.23805 (Sheriff, Huang, Nemeth, Nakhjiri, 2025)
- "Proactive Security for Large-Scale AI Inference: Adaptive Container Orchestration at the Edge" (Sheriff, Nemeth, Huang, 2025)
- "Adaptive Security for AI Infrastructure with Automated Moving Target Defense" (Nemeth, 2025)
- "Phoenix: Adaptive Containers — The Mathematical Imperative for Proactive Defense, Rooted in Game Theory" (R6 Security, 2025)

## Open Source

Phoenix Core is fully open source under an open-source license. Repository: https://github.com/r6security

## Links

- Website: https://r6security.com
- Products: https://r6security.com/products/kubernetes, https://r6security.com/products/ai
- Research: https://r6security.com/research
- Open Source: https://r6security.com/open-source
- Case Studies: https://r6security.com/case-studies
- GitHub: https://github.com/r6security
- arXiv Paper: https://arxiv.org/abs/2505.23805
- Contact: contact@r6security.com