Which ethical principle should guide AI development in the UK?

No AI system should be able to make consequential decisions — from benefit assessments to criminal sentencing — without meaningful human review. Accountability must rest with people, not algorithms.

If an AI system affects your life, you should be entitled to a clear explanation of why. Black-box decision-making erodes public trust and undermines democratic accountability.

Borrowed from medical ethics, this principle insists that the first obligation of any AI system is to avoid causing damage — physical, psychological, or social — to individuals and communities.

Biased training data has already led to discriminatory outcomes in hiring tools and facial recognition. Ensuring equitable treatment across race, gender, and class must be a foundational constraint.

Short-term commercial gains should never override the broader social good. AI policy must prioritise outcomes for the many — not just the tech firms and investors who profit from deployment.