AI Testing and Audit

Independent eyes on your AI, before the market finds the flaws

We beta test and audit AI systems as products: AI models and agents, humanoid and functional robots, AI-enabled drones, and any device or service with an embedded AI system. The question is always the same: does it behave as claimed, fail safely, and stand up to the scrutiny of regulators, insurers and buyers?

What we test

  • Model and agent behaviour: capability claims, guardrails, misuse paths and failure modes.
  • Robots and physical AI: functional testing of humanoid and task robots against their stated envelope, with structured incident and anomaly reporting.
  • AI drones: flight-context testing by an operating drone practice, not a desk review.
  • Embedded AI products: assessment of AI components inside consumer and industrial products, aligned with emerging regulatory expectations for market surveillance and the EU Cyber Resilience Act.
High-tech command centre with digital displays and control panels. People working in a control room with multiple computer monitors.

Credentials

  • ForHumanity Certified Auditor.
  • Working-group member, IEEE P7007.1, the ontological standard for addressing risks in AI systems.
  • Long-standing member of UNECE Working Party 6 groups on risk management and market surveillance in regulatory systems.
  • Hands-on drone operation through our own aerial services practice.

Pricing

By proposal; structured beta programmes from £3,500.

Discuss a testing programme