No. 69 of 500 No. 19 in AI Ethics & Society

Invisible Women

Data Bias in a World Designed for Men

By Caroline Criado Perez

Published
March 2, 2021
Print length
448 pages
Reading level
Beginner
ISBN-13
978-1419735219
Publisher
Harry N. Abrams
Written for
Product designers, data scientists, policymakers and health professionals who need to understand how data gaps create bias.

Synopsis

Caroline Criado Perez argues that much of the modern world runs on data that treats the male body and the male life as the norm, and that the resulting gaps go unnoticed precisely because they never appear in the numbers. Drawing on a large body of studies, she moves through daily life, the workplace, product design, medicine, public life and disaster response. Her examples range from snow-clearing schedules that prioritized commuter roads over the walking routes women use more often, to car crash-test dummies modeled on the average male, to drug trials and diagnostic criteria developed largely on men, to voice recognition software that performs worse on female voices.

For readers interested in AI, the book matters because machine learning systems inherit every gap in their training data. Criado Perez shows how unrecorded unpaid care work, statistics that are not broken down by sex and male-default design choices carry through into algorithms, hiring tools and health care, and she argues that the remedy begins with collecting sex-disaggregated data and bringing women into design decisions. The prose is direct and often pointed, and extensive endnotes let readers check her sources. Readers gain a durable habit of mind: asking who was counted, who was missed and what a dataset assumes before trusting any system that depends on that data.

Scorecard

Editorial score

How it scored

Editorial score: 75.0 / 1005 weighted criteria

  • Influence30%

    25.5/30
  • Authority25%

    17.5/25
  • Practical value20%

    13.0/20
  • Currency15%

    10.5/15
  • Readability10%

    8.5/10
  • Editorial score

    75.0/100

How the rank was computed

Rank No. 69 of 500 comes from a composite of 65.9: observable signals 32.1 of 55 points plus editorial review 33.8 of 45 points.

All 11 inputs
InputPoints
Observable signals32.1/55
Reach (Amazon ratings volume)15.1/16.5
Consensus across best-of lists2.8/13.8
Awards and bestseller evidence5.0/8.3
Reader rating, sample-adjusted6.9/8.3
Amazon sales rank2.3/5.5
Recency0.0/2.8
Editorial review33.8/45
Author authority7.9/11.3
Influence11.5/13.5
Practical value5.9/9.0
Readability3.8/4.5
Currency4.7/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

An exhaustively sourced account of the gender data gap, showing how male-default data yields products and policies that fail women.

Strengths

  • Thoroughly documented with extensive endnotes
  • Concrete examples across medicine, design and policy
  • Directly relevant to training-data bias debates

Keep in mind

It is a broad social survey rather than an AI book, and some readers find its tone polemical or dispute specific statistics and interpretations.

Key takeaways

  1. Data that treats men as the default is not neutral; missing sex-disaggregated data quietly produces products, policies and medical care that fit women less well.
  2. Machine learning systems trained on male-skewed datasets, such as speech recognition and hiring algorithms, amplify existing gaps rather than correcting them.
  3. Medical research that historically underrepresented women contributes to misdiagnosis, including heart attack symptoms that differ from the textbook male presentation.
  4. Closing the gap starts with collecting data disaggregated by sex and including women in the rooms where products, cities and policies are designed.

Notable facts

About the author

Caroline Criado Perez

Caroline Criado Perez is a British writer, broadcaster and campaigner who studied English at Keble College, Oxford. Her campaigns put Jane Austen on the Bank of England £10 note and secured the Millicent Fawcett statue in Parliament Square, unveiled in 2018. She was appointed OBE in 2015 for services to equality and diversity. Her earlier book is Do It Like a Woman.

Questions

What is Invisible Women about?
Invisible Women documents the gender data gap: the missing or male-default data on bodies, work and lives across medicine, product design, transport, the workplace and public policy. Caroline Criado Perez argues that when data treats men as the norm, systems built from it, from car safety testing to drug dosing to voice software, serve women less well, sometimes with serious consequences for health and safety.
Why does Invisible Women appear on a list of AI books?
Although it covers far more than technology, the book gives one of the clearest accounts of how biased inputs produce biased outputs. Its examples from speech recognition, hiring algorithms and medical research show how machine learning can absorb and scale existing gaps. Anyone who builds, buys or audits AI systems gains a practical test: establish what a dataset leaves out before relying on a model trained with that data.
Is Invisible Women worth reading in 2026?
Yes. Its examples predate the generative AI boom, but the problem it describes, unrepresentative data feeding automated decisions, has grown more pressing as models spread into health care, hiring and public services. The book's two major 2019 prizes reflect its standing, and its extensive endnotes let readers check individual figures against the original studies before applying them.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.