AI Design Principles
The foundational principles for designing AI-powered product experiences that are useful, trustworthy, and genuinely human-centred.
Value & Fit
- Familiarity and Learnability
Anchor AI in familiar UX patterns so people can understand and adopt it quickly.
Value & Fit - User Value First
Use AI only where it creates clear user value rather than novelty.
Value & Fit - Task and Context Fit
Use AI only in tasks and situations where it fits the job, risk, and user context.
Value & Fit
Clarity & Trust
Control & Oversight
Safety & Responsibility
- Privacy, Consent, and Permission Awareness
Make data use clear, permission-aware, and respectful of user boundaries.
Safety & Responsibility - Safety, Graceful Failure, and Recovery
Design AI to fail safely, recover clearly, and always give users a way out.
Safety & Responsibility - Fairness and Bias Mitigation
Reduce harmful bias and design for fair outcomes across different people and contexts.
Safety & Responsibility
Data & System Quality
- Performance and Responsiveness
Keep AI fast, visible, and well-paced with progress cues and partial results where possible.
Data & System Quality - Operational Robustness
Ensure the system behaves reliably across real-world conditions, failures, and edge cases.
Data & System Quality - Data Quality and Maintenance
Use relevant, accurate, current data and plan for drift, staleness, and upkeep.
Data & System Quality