Sign in
Keeping humans in charge
Bitesize

Keeping humans in charge

AI governance and oversight in public services.

Level
Bitesize
Learning time
1 hour 15 minutes
Where
AI for Leaders
Price
£39

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

Go to the courseStart the free Taster

Certificate of Completion

About this course

Overview and Outcomes.

Overview

Oversight you cannot see is oversight you do not have. This module separates hollow from meaningful human oversight, insists on a named owner for every AI system, and sets out five building blocks: a register, rules for use, impact assessment, monitoring, and challenge with openness. It applies the three lines model and NIST's Govern, Map, Measure, Manage cycle, reviews standards boards hear about, and describes four board postures. When systems fail, people pay. This module studies Robodebt in Australia, the Dutch childcare benefits scandal, the Post Office Horizon scandal, the 2020 grading algorithm and the DWP model for Universal Credit advances. Comparing them side by side, it draws out five recurring patterns, from warnings ignored to accountability diffused, and argues that efficiency which harms people is not efficient.

What you'll develop

In this Bitesize, you will develop the ability to:

  1. Distinguish hollow from meaningful human oversight and assess it using override rates and what reviewers say.
  2. Describe the five building blocks of AI governance and apply the three lines model and the NIST cycle to AI.
  3. Assess where your board sits among four postures and ask for evidence rather than reassurance.
  4. Describe what went wrong in Robodebt, the Toeslagenaffaire and Horizon, and who bore the harm.
  5. Identify five recurring patterns behind algorithmic failure in public service and recognise their early warning signs.
What it covers

The Modules.

1 hour 15 minutes of learning.

  1. 10Governance in Practice: Oversight That WorksOversight you cannot see is oversight you do not have. This module separates hollow from meaningful human oversight, insists on a named owner for every AI system, and sets out five building blocks: a register, rules for use, impact assessment, monitoring, and challenge with openness. It applies the three lines model and NIST's Govern, Map, Measure, Manage cycle, reviews standards boards hear about, and describes four board postures.37 min
  2. 11When Algorithms Fail: Lessons from Public ServiceWhen systems fail, people pay. This module studies Robodebt in Australia, the Dutch childcare benefits scandal, the Post Office Horizon scandal, the 2020 grading algorithm and the DWP model for Universal Credit advances. Comparing them side by side, it draws out five recurring patterns, from warnings ignored to accountability diffused, and argues that efficiency which harms people is not efficient.38 min
Where it fits

Keep going.

Part of the AI for Leaders topic, also sold whole as The human edge.

How credit works

Also Inside