Clear heads on AI

Who It's For

Leaders deciding how AI is used in their organisation

Overview

AI decisions are reaching leaders' desks faster than most organisations can form a view. Clear heads on AI treats them as leadership questions rather than IT projects. The day moves from how large language models actually work and what the best evidence on productivity shows, through the UK, Welsh and European policy and legal landscape, to fairness, bias and oversight that genuinely works, with lessons from Robodebt, the Dutch childcare benefits scandal and Horizon. It then turns to human judgement, drawing on Kahneman, Simon's bounded rationality and prospect theory, and closes with leading staff through adoption and building a culture where new ideas can land safely.

Aims

This programme aims to give leaders the clarity and confidence to own decisions about AI in their organisation: grounded in evidence rather than enthusiasm, lawful and fair by design, and centred on the residents and staff affected. It develops the judgement to ask better questions, to recognise bias in systems and in ourselves, and to keep meaningful human oversight in place.

In summary, you will be able to:

1. 1Explain how AI tools produce their outputs, what the evidence shows about their value, and why existing legal and ethical duties still apply.
2. 2Assess an AI proposal against the policy landscape, data protection and equality law, fairness risks and the lessons of public service algorithm failures.
3. 3Lead colleagues through AI adoption with named ownership, meaningful oversight and decision-making that guards against bias, overconfidence and shadow use.

What you'll develop

By the end of this Immerse, you will be able to:

1. 1Explain why AI is a leadership question rather than an IT project, and why existing duties of fairness, lawfulness and accountability still apply.
2. 2Explain what the main productivity studies found, including who gained most and where AI made capable people worse.
3. 3Describe the main milestones in UK and Welsh AI policy and the distinct Welsh frame of AI Cymru and social partnership.
4. 4Identify the five routes by which bias enters an AI system and explain why bias scales once it is encoded.
5. 5Describe what AI literacy means at four levels: everyone, managers, specialists, and leaders and boards.
6. 6Distinguish System 1 from System 2 thinking and identify when a strategic question is being answered by the wrong one.
7. 7Describe the five discovery skills of associating, questioning, observing, experimenting and networking, and how a leader can model each one.
8. 8Explain loss aversion, reference dependence and diminishing sensitivity, and how they depart from the classical rational model.