OUR COMMITMENT
Trustworthy AI Policy
The principles we build by, and the ones our platform helps you apply.
Our view on AI
AI is changing how organisations decide, serve customers and grow. Used well, it creates enormous value. Used without oversight, it becomes an opaque liability that no one can explain, correct or defend.
We believe AI should be visible, explainable and accountable to the people it affects, and that good governance is what makes this possible at scale.
Our seven principles
1. Human oversight
People stay in charge. AI supports decisions; accountable humans make them, and can step in at any time.
2. Technical robustness
AI systems should be accurate, reliable and secure, and monitored so problems are caught early.
3. Accountability
Every AI system has a named owner, and every decision about it is recorded.
4. Transparency
It should always be possible to see what an AI system does, what data it uses and why it produced a result.
5. Human-centred design
AI should serve people's needs and respect their rights, not the other way round.
6. Fairness and societal impact
We consider who an AI system could affect, and work to prevent unfair or harmful outcomes.
7. Legal compliance
AI must meet the laws and regulations that apply to it, including the EU AI Act.
Principles in practice
Principles only matter if they shape real decisions. We use these principles to assess how we build our own platform and the AI we use internally. Sometimes that means accepting trade-offs, for example between full transparency and protecting sensitive data. And sometimes the most responsible choice is not to build something at all.
How our principles relate to others
Our principles draw on widely recognised frameworks, including the EU AI Act, the OECD AI Principles, the EU Ethics Guidelines for Trustworthy AI and ISO/IEC 42001. What sets our approach apart is a focus on putting these principles into daily practice through the platform, so they are enforced, not just written down.
- EU AI Act
- OECD AI Principles
- EU Ethics Guidelines for Trustworthy AI
- ISO/IEC 42001