As RetailFlow delivers its four funded AI initiatives (customer-service chatbot, dynamic pricing, inventory optimisation and fraud detection), we commit to building and using AI in a way that is consistent with our values, particularly Integrity and Customer First. These principles apply to every AI system we build, buy or operate.
1. Human oversight
AI supports our people; it does not replace their judgement on decisions that materially affect customers or staff. Every AI system must have a clearly accountable human owner, a documented escalation path, and the ability for a person to review, override or switch off automated decisions. High-impact actions (e.g. fraud blocks, large price moves) require a human-in-the-loop check.
2. Data privacy & sovereignty
We collect and use only the data we need, with a lawful basis, in line with the Australian Privacy Act and the Australian Privacy Principles. Customer data is stored and processed in Australian data regions wherever practicable, and we will not transfer personal data offshore without a documented assessment. Vendors must meet equivalent standards.
3. Transparency
We are open about when customers are interacting with AI. The customer-service chatbot will clearly identify itself as automated and offer a path to a human. Internally, we document how each model is trained, what data it uses, and the limits of its reliability.
4. Fairness
Our AI must not create unfair or discriminatory outcomes. We will test models for bias across customer segments before launch and monitor them in production. Dynamic pricing must not exploit vulnerable customers or produce outcomes that breach the Australian Consumer Law.
Accountability
Proposed governance: an AI Review Group chaired by the CIO (Marcus Kim), with the GM Customer Experience (Lisa Nguyen), CFO (David Chen) and Legal, signing off each initiative before go-live and reviewing performance quarterly.
Drafted by Dr. Priya Sharma, Head of Data & Analytics, with Legal