Best books for AI Ethics Officers

Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy

by Cathy O'Neil · 2016

~270 pages~7 hrs to read

Case after case of scoring and prediction models that harm people while looking objective, written by a former quant. A good first read for seeing what an AI ethics review is actually looking for.

Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence

by Kate Crawford · 2021

~290 pages~7 hrs to read

Follows AI systems back to their material inputs: mining, data labelling, energy and the institutions that fund them. Read it early to widen the question from 'is this model fair' to 'who bears the cost of this system'.

The Ethical Algorithm: The Science of Socially Aware Algorithm Design

by Michael Kearns, Aaron Roth · 2019

~260 pages~7 hrs to read

Two computer scientists explain how fairness, privacy and accountability can be written into algorithms, and where the trade-offs sit. Accessible enough for a non-engineer who needs to challenge a technical team.

Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor

by Virginia Eubanks · 2018

~290 pages~7 hrs to read

Reported studies of automated eligibility, risk-scoring and policing systems in US public services. Shows what harm looks like when it lands on real people, which is useful context for any impact assessment.

Fairness and Machine Learning: Limitations and Opportunities

by Solon Barocas, Moritz Hardt, Arvind Narayanan · 2023

The standard textbook on how fairness is defined, measured and contested in machine learning. Read it when you start running bias audits and need to understand what each metric can and cannot tell you.

Practical Fairness: Achieving Fair and Secure Data Models

by Aileen Nielsen · 2020

~280 pages~7 hrs to read

A hands-on guide to checking datasets and models for unfair outcomes, with the legal and security angles alongside the code. Useful when you are working directly with data science teams on audits.

Machine Learning for High-Risk Applications: Approaches to Responsible AI

by Patrick Hall, James Curtis, Parul Pandey · 2023

Covers governance, model risk management, explainability, bias testing and security for ML used in consequential settings. Written for practitioners who have to put controls around a real system.

Responsible AI: Best Practices for Creating Trustworthy AI Systems

by Qinghua Lu, Liming Zhu, Jon Whittle, Xiwei Xu · 2024

Turns responsible-AI principles into practices across the lifecycle, from governance and process to product-level design patterns. Helpful when you are building an internal framework rather than just reviewing models.

The Alignment Problem: Machine Learning and Human Values

by Brian Christian · 2020

~360 pages~9 hrs to read

A reported history of how researchers have tried to make machine learning systems do what we actually intend, covering fairness, transparency and reinforcement. Read it when you advise leadership and need the field's full picture.

Human Compatible: Artificial Intelligence and the Problem of Control

by Stuart Russell · 2019

~340 pages~9 hrs to read

An AI researcher argues for building systems that stay uncertain about human preferences and under human control. Relevant when your remit reaches long-term risk and board-level questions.

Unmasking AI: My Mission to Protect What Is Human in a World of Machines

by Joy Buolamwini · 2023

~340 pages~9 hrs to read

A first-hand account of finding bias in facial-analysis systems and pushing for accountability, told by the researcher who did the work. A model for how to make a case that organisations act on.

Race After Technology: Abolitionist Tools for the New Jim Code

by Ruha Benjamin · 2019

~170 pages~4 hrs to read

Examines how technology can reproduce racial hierarchy under a neutral surface, and what designers and organisations can do about it. Best for seasoned reviewers who want to sharpen how they question 'neutral' systems.