Platform

Evidence-Bound Assurance
for Consequential AI

HAIEC connects source, identity, policy, effective authority, code capability, runtime evidence, and observed effects to show what an AI system can cause — and exactly what the available evidence can support.

Run the same Assurance workflow from the HAIEC workspace, a connected GitHub repository, the Assurance API, or an MCP-enabled AI agent.

Explore the Kestrel System Constellation Book Demo

HAIEC complements existing security, IAM, observability, CI/CD, and governance tooling. Where supported, their evidence can contribute to a HAIEC Evaluation. HAIEC does not replace those systems or convert their findings directly into Assurance truth.

How HAIEC Works

Define → Connect → Collect Evidence → Assure → Verify → Monitor

See the System

The AI Action & Access Map shows how identities, permissions, AI actions, APIs, data, and consequences connect.

AI execution → action surface → consequence
Credential evidence and access paths
What controls apply

Explore a sample map

1. Collect Evidence

Source, identity, access, policy, capability, and runtime evidence where supported.

Source code analysis (AI AppSec)
AI inventory and identity evidence
Runtime observation where connected

2. Compare

Five evidence questions: what you intend, what policy allows, what credential evidence establishes, what the application can do, and what was observed.

Requested vs. authorized vs. granted
Code-capable vs. observed
Unknown and partial made explicit

3. Assure

Bounded evaluation with explicit limitations. ALLOW means ALLOW within evaluated scope — not “safe” or “certified.”

ALLOW / REVIEW / BLOCK evaluation
Unknown never silently becomes PASS
Coverage and limitations stated

Verify

Assurance packages and decision receipts with tamper-evident integrity and traceable provenance where supported.

Evidence packages with provenance
Tamper-evident integrity
Reviewer handoff and export

Monitor

Re-evaluate as systems, credentials, and code change. Continuous monitoring where connected — not universally claimed.

CI/CD integration and GitHub App
Re-scan on change
Evidence refresh where supported

Capabilities

Evidence and security tools that feed the assurance lifecycle

AI Security

Deterministic Source Analysis

Source analysis for AI-specific security findings, action surfaces, capability paths, and code-capable evidence.

Runtime Security

Authorized Runtime Testing

Controlled runtime testing against supported endpoints after explicit authorization. Runtime results become evidence; they do not independently determine Assurance.

Compliance Twin

Evaluation Records

Versioned assurance evaluation records with audit trails for compliance history tracking.

Evidence Generator

Automated Artifacts

Generate audit-ready evidence packages for SOC 2, ISO 27001, GDPR, and more frameworks.

Compliance Wizard

Guided Assessments

Step-by-step compliance assessments for SOC 2, ISO 27001, EU AI Act, NYC LL144, and more.

AI Inventory

Discovery Engine

Automatically discover and catalog all AI systems, models, and data sources across your organization.

GitHub Integration

CI/CD Native

Auto-scan pull requests, flag code that may not meet your policies, and generate compliance reports in your workflow.

REST API

Programmatic Access

Full API access for custom integrations, automated workflows, and enterprise deployments.

Open Source SDKs

npm & PyPI

Runtime monitoring packages for Node.js and Python. Install, instrument, and monitor in minutes.

LLMverify

Free Tool

Free online tool to test your prompts for injection attacks, jailbreaks, and security vulnerabilities.

AI Exposure Score

Free Assessment

25-question assessment to measure your AI compliance readiness across 5 critical domains.

Bias Detector

NYC LL144 Bias Audit Support

Detect and measure bias in AI hiring tools. NYC Local Law 144 bias audit support.

Why Teams Choose HAIEC

See how we compare to traditional approaches

Feature	HAIEC	OneTrust	Manual
AI Security Scanning	Built-in	Add-on	N/A
Evidence Collection	Automatic	Manual	Spreadsheets
GitHub Integration	Native	None	N/A
Pricing	$0-$999/mo	$50k+/year	Staff cost
Setup Time	5 minutes	3-6 months	Ongoing
Open Source SDKs	npm/PyPI	Closed	N/A

Ready to Build Defensible Assurance?

Choose your path to get started

For Developers

Start with a free scan. No signup required.

Scan a Repository

GitHub Integration

Scan repos on every PR. Free to install.

Install Free GitHub App

For Compliance Teams

See how HAIEC fits your compliance needs.

Or explore:Sample Reports Documentation Pricing