AI Ethics Statement

Prodia commits to the ethical development and operation of autonomous software engineering systems. The following principles are reflected in product design, engineering practice and contractual commitments.

1. Fairness

• Treat similarly situated customers, repositories and users consistently.
• Identify and mitigate sources of systematic bias in evaluations, prioritisation and routing.
• Document and review residual risk where bias cannot be fully eliminated.

2. Transparency

• Every autonomous action carries an explainability record describing intent, inputs, evidence considered, and confidence.
• Material limitations and known failure modes are disclosed in the AI Safety Statement and product documentation.
• Where AI is used in customer-facing interactions, that fact is disclosed under the AI Transparency Statement.

3. Accountability

• Prodia maintains a single, identifiable internal owner for every governance control and every customer-facing safety surface.
• Audit trails are designed to support after-the-fact accountability, including by regulators and independent reviewers.
• Operational responsibility for production systems remains with the customer, as set out in the Autonomous SDLC Governance statement.

4. Human control

• No production effect without human-configured authority.
• Human approvers may always override, defer or roll back proposed actions.
• Customers may reduce, restrict or revoke autonomy at any time.

5. Safety

• Reversibility, sandboxing, falsification and rollback are first-class engineering requirements, not afterthoughts.
• Foreseeable misuse is considered in design review and addressed in the Acceptable Use Policy.

6. Out-of-scope uses

Prodia is not designed for, and may not be used for, the development of systems whose primary purpose is to cause harm, to manipulate persons in ways prohibited by the EU AI Act, to enable mass surveillance contrary to fundamental rights, or to circumvent applicable law.