AI Policy

1. AI in Development & Testing

We utilize Artificial Intelligence and Large Language Models (LLMs) as productivity tools during the development phase. This includes:

• Generating boilerplate code and refactoring existing patterns.
• Enhancing security testing by identifying potential vulnerabilities in pre-production code.
• Accelerating documentation and unit test generation.

All AI-generated code is subject to the same rigorous human peer-review process as traditionally written code before it is integrated into our platform.

2. The "Air Gap" for Production

We maintain a strict "Air Gap" policy between AI tools and our production environment. AI models never touch production infrastructure. Our deployment pipelines, server configurations, and live operational systems are managed through deterministic, human-controlled processes.

3. Zero Customer Data Exposure

Protecting your information is our highest priority. AI tools never have access to customer data. We do not use customer databases, case files, or any sensitive operational data to train models or to provide context to third-party AI services.

4. Model Selection & Open Source

Whenever possible, we prefer using open-source or "open-weight" models. This allows us to audit the models more effectively and ensures we are not locked into proprietary ecosystems that may have opaque data handling practices.

5. Infrastructure & Sovereignty

In alignment with our Sovereign Data Policy, we aim to run AI models on infrastructure we control or on Canadian-hosted servers. This ensures that even the tools we use to build our software remain under appropriate jurisdictional oversight.

6. Continuous Oversight

As AI technology evolves, so will our safeguards. We continuously monitor developments in AI safety and privacy to ensure our practices remain at the forefront of responsible software engineering.

Last Updated: April 29, 2026