ZeroTrusted AI Firewall
ZeroTrusted AI Firewall inspects prompts, model responses, and tool calls to help organizations control AI use.
What ZeroTrusted AI Firewall is
ZeroTrusted AI Firewall inspects prompts, model responses, and tool calls to help organizations control AI use. It can block prompt injection, jailbreaks, data exfiltration, hallucinations, toxicity, and unauthorized tool use. Before prompts leave an organization’s perimeter, it can redact or block personal and health information, secrets, classified markings, and custom-sensitivity tokens. Administrators can allow-list agent tools, review arguments, set argument types, and revoke tool access after a policy violation. Deployment choices include inline proxy enforcement, out-of-band traffic mirroring for detection and tuning, and browser or edge enforcement when paired with Shadow AI Protection. It can sit between users and ChatGPT, Claude, Gemini, self-hosted models, or internal agent frameworks. The maker says inline deployment adds under a millisecond to most inference paths; out-of-band monitoring has no production-path latency impact. Pricing is custom and available on request, with a 15-day trial. The Enterprise tier includes SaaS, on-premises, air-gapped, and multi-cloud options.
Good
- Redacts or blocks sensitive data before prompts leave the perimeter
- Supports tool allow-lists and argument controls
- Offers inline, mirrored, browser, and edge enforcement options
- Lists ChatGPT, Claude, Gemini, and self-hosted models
- 15-day trial is available
Watch out
- No free plan
- Pricing is custom and requires contacting sales
- Enterprise tier is the only tier with listed onboarding and dedicated success manager
Who it suits
It suits organizations that need to govern AI prompts, outputs, and agent tools, especially where sensitive information or compliance requirements matter. Deployment options cover production enforcement as well as monitoring and policy tuning.
Verdict
ZeroTrusted AI Firewall combines prompt and output inspection with sensitive-data controls and agent-tool governance. Its custom pricing and organization-focused deployment options make it a product to evaluate with sales rather than a self-serve firewall.
Plans and prices
From ZeroTrusted AI Firewall's own pricing page, checked 30 Sep 2026.
| Plan | Price | What it gives you |
|---|---|---|
| Enterprise | Priced on request | Custom pricing tailored to security, compliance, and deployment needs · SaaS, on-premises, air-gapped, and multi-cloud · volume and federal procurement options |
The facts
Each one read from ZeroTrusted AI Firewall's own pages, with the date it was checked.
What it does
- API protection
- It can rate-limit, throttle, and isolate abusive callers at the AI API layer, with adaptive rules based on live traffic patterns.checked 30 Sep 2026
- Data protection
- It redacts or blocks PII, PHI, secrets, classified markings, and custom-sensitivity tokens before prompts leave an organization’s perimeter.checked 30 Sep 2026
- Deployment
- The listed modes are inline proxy for production-path enforcement, out-of-band traffic mirroring for detection and policy tuning, and browser or edge enforcement when combined with Shadow AI Protection.checked 30 Sep 2026
- File sanitization
- The dashboard can sanitize uploads and prompts by redacting PII, PHI, classified markings, and secrets; the page names English, Japanese, Spanish, Portuguese, and more.checked 30 Sep 2026
- Latency
- The maker says inline deployment adds under a millisecond to most inference paths and out-of-band monitoring has zero production-path latency impact.checked 30 Sep 2026
- Model compatibility
- The product page names ChatGPT, Claude, Gemini, self-hosted LLMs, and internal agent frameworks as models or environments it can sit between users and.checked 30 Sep 2026
- Pricing limits
- Pricing is custom and the maker directs organizations to contact sales; the page lists a dedicated success manager and onboarding with the Enterprise tier.checked 30 Sep 2026
- Privacy
- The privacy policy says data is encrypted in transit and at rest, the company does not sell personal data, and it does not use third-party tracking cookies or analytics scripts.checked 30 Sep 2026
- Purpose
- AI Firewall inspects prompts, model responses, and tool calls to block prompt injection, jailbreaks, data exfiltration, hallucinations, toxicity, and unauthorized tool use.checked 30 Sep 2026
- Security and compliance
- The company’s About page says its platform has SOC 2, ISO 27001, NIST, and HIPAA compliance built in.checked 30 Sep 2026
- Support
- The contact page invites visitors to request a demo or security assessment and says communications are encrypted end-to-end.checked 30 Sep 2026
- Threat detection
- The product describes adversarial ML defenses for model inversion, membership inference, and poisoning, plus zero-day LLM threat detection through behavioral analytics.checked 30 Sep 2026
- Tool governance
- Administrators can allow-list agent tools, inspect arguments, enforce argument types, and revoke tool access when a policy violation occurs.checked 30 Sep 2026
The company
- Headquarters
- Kissimmee, Florida, United Stateschecked 28 Sep 2026
Alternatives to ZeroTrusted AI Firewall
- 1
PromptGuard
Free planWebAPIExtensionLinux7.9From $19 a month - 2
Prompt Inspector
Free planOpen sourceWebAPILinuxmacOS7.6Free plan - 3
CyberSecEval
Free planOpen sourceWebAPILinuxmacOS7.5Free plan - 4
NVIDIA NeMo Guardrails
Free planOpen sourceAPILinuxmacOSSelf-hosted7.4Free plan - 5
Openlayer Guardrails
Free planWebAPISelf-hosted6.9Free plan - 6
GuardionAI
WebAPILinuxmacOS6.8