Holding AI to account
AI governance: data, law and the EU AI Act.
A focused around four hours on holding AI to account: data, procurement, the law, the EU AI Act, fairness and oversight that works.
- Level
- Focus
- Learning time
- 4 hours
- Where
- AI for Leaders
- Price
- £69
When you step up to a higher level, what you paid in the last 12 months for lower-level courses inside the bigger course comes off its price, in proportion to how much of each is inside it, whichever topic it's from. Courses at the same or a lower level don't give credit against each other. A Focus course or above always costs at least £15. Read the full terms
Certificate of Completion, or Certificate of Completion and Assessment with a short assessment
Overview and Outcomes.
Who It's For
Leaders, board members and senior officers overseeing AI in public services
Overview
Once AI is in use, the hard questions are about accountability. This programme follows them from the data beneath a system, through procurement and contracts, to the law, fairness and oversight. It draws on the Data (Use and Access) Act 2025, the EU AI Act's four risk tiers, Kleinberg and Chouldechova on why fairness has competing meanings, and the three lines model and NIST cycle for governance. It closes by studying Robodebt, the Dutch childcare benefits scandal and Horizon, drawing out the recurring patterns that turn a system failure into human harm.
Aims
The aim is to equip leaders, board members and senior officers to hold AI to account in practice: asking the right questions of data, suppliers and systems, recognising where law and fairness demand stronger safeguards, and building oversight that is meaningful rather than a rubber stamp, so that accountability stays with the organisation.
In summary, you will be able to:
- Assess whether an AI system rests on sound data, a robust contract and lawful foundations, using tests of data quality, supplier due diligence and existing law.
- Distinguish competing meanings of fairness and apply an ethics check that asks for performance by group before a system goes live.
- Design oversight that works, with named owners, a register, monitoring and board challenge informed by the recurring patterns behind public service algorithm failures.
What you'll develop
Through this Focus course, you will build the ability to:
- Explain why AI outputs inherit the quality of their data and why named ownership of data sets matters.
- Explain why accountability for an AI system stays with your organisation and choose whether to build, buy, partner or wait.
- Identify the main areas of existing law that apply when AI is used in a public service.
- Identify the five routes by which bias enters an AI system and explain why bias scales once it is encoded.
- Distinguish hollow from meaningful human oversight and assess it using override rates and what reviewers say.
- Describe what went wrong in Robodebt, the Toeslagenaffaire and Horizon, and who bore the harm.
Built from 3 parts.
4 hours of learning.
- 01Good data, wise buyingData readiness and AI procurementBitesize2 modules
- 02Right, fair and lawfulEU AI Act, AI ethics and algorithmic biasBitesize2 modules
- 03Keeping humans in chargeAI governance and oversight in public servicesBitesize2 modules
Go deeper, and pay only the difference
- You are hereFocus, 4 hours, £69
- The whole topicImmerse, The human edge, 11 hours, £249. Already own the Focus? You pay £180.
When you step up to a higher level, what you paid in the last 12 months for lower-level courses inside the bigger course comes off its price, in proportion to how much of each is inside it, whichever topic it's from. Courses at the same or a lower level don't give credit against each other. A Focus course or above always costs at least £15. Prices per person. Read the full terms
Also part of: Leading with data and digital, Digital with people in charge