AI governance, compliance, and security for production AI.
NorthRail AI helps teams assess AI risk, map findings to governance frameworks, and create evidence they can use for release decisions. RedTeam AI is the first app in the platform, focused on automated LLM red-teaming.
Why traditional security fails AI
LLMs break in ways WAFs can't see.
LLMs fail differently
Firewalls and static analysis can't catch prompt injection, data leakage, or jailbreaks. AI systems require a new category of security testing.
Dynamic vulnerabilities
AI responses change with context, temperature, and phrasing. A single test is not enough; you need automated adversarial probing at scale.
Compliance is coming
EU AI Act, NIST AI RMF, and ISO 42001 require documented evidence of AI risk assessment. Manual red-teaming doesn't scale or create audit trails.
How it works
From endpoint to report in minutes.
Define your target
Point RedTeam AI at your AI endpoint, including REST API, SSE stream, or WebSocket. Configure authentication and select your AI architecture type (chatbot, RAG, agent, API).
Run automated scans
Choose a scan preset such as Quick Check, Standard OWASP, Deep Scan, or Custom, or use a domain preset tuned for finance, healthcare, agents, and more. RedTeam AI runs adversarial probes across four specialized engines: vulnerability probing, adversarial prompting, attack orchestration, and response evaluation. No manual scripting required.
Get actionable reports
Findings are auto-triaged by severity with false-positive filtering. Reports map vulnerabilities to EU AI Act, NIST, and ISO 42001. Export audit-ready evidence (DOCX, PDF, JSON, SARIF), share with approvers, or wire into CI/CD to gate releases on scan results.
Who it's for
Built for technical teams.
AI/ML Teams
Test your models before deployment. Find prompt injection, data leakage, and jailbreak vulnerabilities before they reach production.
Security Engineers
Extend your security program to AI systems. Run repeatable adversarial tests, export SARIF, and turn findings into release decisions.
Compliance Officers
Get reusable evidence for customer reviews, EU AI Act, NIST AI RMF, ISO 42001, and internal governance records.
Pricing
Simple, transparent pricing.
Start free with 2 scans every month. No credit card required.
Free
- ✓ 2 scans per month
- ✓ OWASP LLM Top 10 preset
- ✓ All findings on screen
- ✓ Directional mitigation on every finding
- ✓ Full remediation preview on top finding
Starter
- ✓ 10 scans per month
- ✓ All presets (OWASP, Agentic, Custom)
- ✓ Evidence export in DOCX, PDF, JSON, and SARIF
- ✓ Full compliance mapping (EU AI Act, NIST, ISO 42001, OWASP)
- ✓ Basic share link · 90-day retention
Professional
- ✓ 25 scans per month
- ✓ Everything in Starter
- ✓ API access + CI/CD integration
- ✓ Release-gate workflow
- ✓ Advanced share controls · 12-month retention
Team
Coming soon$349/mo · 75 scans/mo · organization controls
SSO, RBAC, team workflows, and 24-month retention are rolling out for larger teams. Get notified when Team launches.
Prefer to pay as you go?
One-time scan packs with no subscription. Every pack unlocks full evidence export and stays valid for 12 months.
Starter Pack
$99
10 scans · $9.90/scan
Standard Pack
$249
30 scans · $8.30/scan
Bulk Pack
$699
100 scans · $6.99/scan
Compliance frameworks
Findings map to AI governance and security frameworks across US, EU, and Brazil regimes, selectable per scan target.
AI architectures supported
Ready to govern and secure your AI?
NorthRail AI brings security testing, compliance mapping, and governance evidence into one platform for production AI teams.
Explore RedTeam AINo credit card required. 2 free scans every month.