Independent Verification, Real Accountability

We exist because AI systems need scrutiny from outside the organisations that build them.

Our team at work

Why We Started This

In 2023, a major Australian retailer deployed a customer service AI that systematically provided worse service to customers from certain postcodes. The company only discovered this after six months and significant reputational damage.

The problem wasn't malicious intent. It was the absence of external verification.

Internal teams rarely spot their own blind spots. They're too close to the technology, too invested in its success, too constrained by corporate pressures to ship quickly.

That retailer needed someone who could evaluate their AI objectively, without corporate allegiances or product timelines influencing the analysis.

We became that someone.

Our Verification Process

We don't just review documentation. We test live systems against adversarial scenarios, edge cases, and real-world conditions that expose hidden failures.

System Mapping

Understanding data flows, decision points, and dependencies within your AI architecture.

Adversarial Testing

Probing the model with inputs designed to reveal bias, errors, and unexpected behaviour.

Fairness Analysis

Measuring outcomes across demographic groups to identify discriminatory patterns.

Compliance Review

Verifying alignment with Australian privacy laws, industry standards, and ethical guidelines.

Remediation Planning

Providing actionable recommendations with implementation priorities and risk assessments.

What Guides Our Work

Independence

We never consult for AI vendors whose products we might later audit. No conflicts, no compromises.

Rigour

Our assessments follow documented methodologies with reproducible results. Every claim we make is backed by evidence.

Practicality

We understand business constraints. Our recommendations balance risk mitigation with operational reality.

Transparency

You receive complete reports detailing our testing approach, findings, and reasoning. No black-box assessments.

Who Does This Work

Our team combines technical depth with domain expertise across law, ethics, and industry-specific knowledge.

AI researchers who've published on fairness and robustness. Privacy lawyers who understand Australian regulatory nuances. Industry specialists from finance, healthcare, and government sectors.

We're not generalists offering AI audits as a side service. This is all we do.

What Makes Our Approach Different

We Test Behaviour, Not Documentation

Most audits verify that policies exist. We verify that systems actually behave according to those policies under real conditions.

We Hunt for Failures

We're incentivised to find problems before they become public. That means stress-testing your AI in ways your own team hasn't considered.

We Speak Multiple Languages

Technical findings mean nothing if executives can't understand them. We translate AI risks into business language that enables decision-making.

Ready to Examine Your AI Systems?

Independent verification starts with a conversation about your specific risks and requirements.

Get in Touch