How we use AI
Our role is not to predict where AI is heading.
We focus on where it can add value today, and apply it responsibly.
People-driven
AI is changing security testing. We’re approaching it with discipline.
We use AI as a tool that complements our experts’ skills and knowledge. It helps our team work faster and cover more ground, without replacing the judgement a specialist brings to your environment. Every finding is verified and signed off by a person before it reaches you.
Data sovereignty
Flexible to your comfort level
Different organisations have different appetites for where their data goes. We offer three options, and you choose one before an engagement starts.
Frontier models
AI tools powered by Anthropic’s Claude and OpenAI’s GPT, wielded by our specialists.
We hold approval for cyber use cases. Your data is handled under our agreements with AI providers.
Your own API keys
Processing stays within your organisation’s existing agreements and terms.
Requires pre-existing approval for cyber use cases from your AI provider.
Local models
Hosted entirely in our office, for organisations with strict data sovereignty requirements.
None of these change how the engagement is scoped, delivered or reported. They only change how your data is handled.