Ethical Product Design

Why ethical design matters

Affordit handles decisions that can feel personal, uncertain and high pressure. Product design should help users understand trade-offs without exploiting that uncertainty.

These principles apply to planning, account creation, subscriptions, commercial journeys, institution-provided access and AI-supported explanations.

Product principles

• Plan first, product second
• No shame, fear or artificial urgency
• No hidden sponsorship or disguised recommendations
• No dark patterns in consent, account deletion or subscriptions
• Clear assumptions, limitations and routes to challenge an output
• Best interests of under-18 users are a primary design consideration
• Institutions receive aggregate reporting by default
• Accessibility and plain language are part of release acceptance

User control

Users should be able to correct inputs, preview a change before applying it, understand why guidance appeared and decide whether to save, share or delete a plan.

Account deletion, consent withdrawal, cancellation and refusal of an optional commercial journey should not be hidden behind misleading wording or unnecessary friction.

Designing for young and vulnerable users

The best interests of users aged 16 and 17 are a primary consideration. Journeys should use age-appropriate language, privacy-protective defaults and proportionate data collection.

High-pressure outcomes should explain uncertainty and practical options without shame, fear or false urgency. Appropriate specialist support should be signposted where Affordit is not the right service.

Commercial boundaries

Commercial relationships cannot alter a calculation, score, verdict, confidence indicator or recommendation ranking. Sponsored or affiliate content must be labelled and kept outside the result.

Affordit should not use false scarcity, disguised advertising, manipulative subscription design or inappropriate targeting of younger users.

AI and automated explanations

AI may explain approved plan context but should not silently change a plan or make a legally significant decision. Deterministic calculation tools remain the source of numerical outputs.

Users should be told where AI is used, be able to challenge an output and receive an explanation of important limitations or missing information.

Governance

A product statement is not evidence that a control operates. Approval requires evidence from the live service or a clearly labelled remediation action and deadline.

Material changes are reviewed against privacy, security, safeguarding, accessibility, financial-boundary and commercial-independence requirements. Every published policy has a named owner and a recorded update date.

How these principles are tested

• Consumer-style journey reviews that identify hesitation and confusion
• Accessibility, keyboard, mobile and reduced-motion acceptance
• Calculation and recommendation validation
• Ownership, privacy and security tests
• Review of commercial labelling and recommendation independence
• Feedback, complaint and incident trends