Subduxion Responsible AI Policy

At Subduxion, we believe AI should amplify human capability, not replace judgment. As a provider of managed AI systems and automations, we build solutions that are explainable, accountable, and aligned with our clients’ business and ethical standards.

Our guiding principles

• Transparency: We make it clear when and how AI is used within our systems. Clients understand the purpose, data inputs, and decision boundaries of each deployed model or agent.

• Human oversight: All AI systems we deploy operate under human governance. We design workflows where people remain in control of key decisions and can override automated outcomes.

• Privacy and data protection: AI systems are trained and operated in compliance with GDPR and applicable data laws. We do not use client data for model training unless explicitly agreed in writing.

• Fairness and non-discrimination: We take measures to detect and mitigate bias in data and models. Our goal is to ensure AI decisions are consistent, fair, and auditable.

• Explainability: Every model and automation is designed with traceability in mind. We prioritize interpretable results and maintain documentation that allows clients to understand why and how outputs are generated.

• Security by design: All AI models, data pipelines, and APIs are protected under the same principles defined in Subduxion’s Information Security Policy - including encryption in transit and at rest, strict access controls, and continuous system monitoring. Security is integrated from design through deployment, ensuring our AI systems meet the same standards of resilience and governance as the rest of our infrastructure.

• Compliance and accountability: We monitor emerging regulations, including the EU AI Act, and adapt our practices accordingly. Clients retain ownership of their data, and Subduxion maintains responsibility for the systems we operate.

• Continuous improvement: We periodically review models, performance, and governance frameworks to ensure our AI remains accurate, relevant, and aligned with ethical standards.

Our promise

We do not deploy or support AI systems designed to deceive, manipulate, or cause harm. Our commitment is to create technology that strengthens human decision-making, promotes transparency, and builds long-term trust between organizations and their data.

For questions or requests regarding Subduxion’s responsible AI practices, contact legal@subduxion.com.