Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsCode Dependent: Living in the Shadow of AI can be read convincingly as a critique of data colonialism, even though Madhumita Murgia does not need to use that exact term for the interpretation to work. Her reported portraits show how human activity becomes data, how low-paid workers build supposedly autonomous systems, and how institutions use classifications designed elsewhere to make decisions about people with little opportunity for appeal.
The book is not a technical history of artificial intelligence or a prediction about superintelligence. It is reported narrative nonfiction about people living with automated systems in work, health, education, migration, identity and political life. That human scale is what makes its larger political argument visible.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Code Dependent: Living in the Shadow of AI | $16.69 | Buy on Amazon |
| 2 |
|
Code Dependent：living in the Shadow of AI (Chinese Edition) | $52.50 | Buy on Amazon |
| 3 |
|
Co-Dependents Anonymous | $22.81 | Buy on Amazon |
| 4 |
|
Codependents' Guide to the Twelve Steps | $9.99 | Buy on Amazon |
What Murgia is arguing about AI
Murgia, whom Macmillan identifies as the Financial Times’ first Artificial Intelligence Editor, follows ordinary people rather than laboratories, founders or investors. The publisher describes cases involving a British poet, a Pittsburgh UberEats courier, an Indian doctor, a Chinese activist in exile, a child assessed as a possible future criminal and a remote community using an AI-assisted diagnostic application. The book asks what happens when systems classify people, predict their behavior, mediate access to work and services, and turn everyday activity into data. Macmillan’s overview presents the stakes as relationships, education, employment, finances, public services, human rights and the weakening of control over decisions made about us.
Those stories are not just a collection of warnings. Read together, they describe a political economy: someone captures the data, someone performs the hidden labor, someone owns the infrastructure and someone else lives with the prediction or decision. “AI” in the book therefore includes predictive scores, biometric identification, algorithmic management, diagnostic tools, content classification and surveillance—not only generative chatbots.
#1 Best Overall
What “data colonialism” means
Data colonialism is the conversion of human life into a continuing source of extractable data, followed by the appropriation of that data for profit, prediction, governance or control. Nick Couldry and Ulises A. Mejias develop the concept as a way to understand how data relations reproduce a colonial logic: social life is treated as a resource that powerful actors can capture and process. Their foundational article is available through SAGE.
The term is related to, but not interchangeable with, several others:
| Term | What it emphasizes |
|---|---|
| Data colonialism | Extraction and appropriation of social life as data. |
| Digital colonialism | Dependence created by platforms, networks, infrastructure, standards and proprietary technologies. |
| Algorithmic colonialism | The export of models, categories and assumptions shaped in one setting into another. |
| Surveillance capitalism | Monetizing behavioral data through prediction and influence. |
| AI supply-chain exploitation | Hidden, often low-paid labor used to label, moderate and evaluate systems. |
Murgia’s cases touch all of these ideas. The data-colonialism lens is most precise when the question is who extracts value, who controls the resulting systems and who bears the consequences.
Ian Koli and the labor behind “automated” intelligence
The clearest entry point is Ian Koli, a Kenyan worker doing data annotation for Sama. In the Macmillan excerpt, his work in Nairobi’s Kibera involves producing detailed labels for datasets used to train systems built for global corporations.
Recommended Free Tools
The apparent contradiction matters. Koli contributes to products marketed as autonomous, yet his own effort is essential and largely invisible to the end user. Annotation can provide formal employment, income, structure and a route to opportunity. It can also be low-paid, psychologically demanding and organized so that much of the commercial value is captured elsewhere. Personal benefit and structural inequality can coexist.
That is why the strongest criticism does not portray every Kenyan data worker as helpless. It asks who sets the rates, owns the resulting datasets and models, allocates risk, and has the bargaining power to change the terms. A 2026 CHI study based on interviews with 18 Kenyan data workers reports dependence, precarity, wage arbitrage and unequal task allocation in the global AI supply chain; it is useful corroborating scholarship, not proof that every worker has the same experience. Read the study.
From lived experience to prediction
Several of Murgia’s subjects encounter systems that turn uncertain possibilities into authoritative-seeming classifications. A child may be assessed as a future criminal; a worker may be ranked by an opaque platform; a patient or service user may be filtered through an automated eligibility or triage process.
The critical questions are concrete:
- What information is collected, and who decided that it was relevant?
- Which categories and historical assumptions shape the model?
- Who is most likely to be misclassified?
- Can the person see, correct or challenge the output?
- What happens when a probability becomes an institutional fact?
The injury is not limited to technical error. A person can lose agency when an institution treats a probabilistic score as a settled description and offers no practical route to contest it. Murgia’s concern is the normalization of decisions that affect a life while remaining difficult to inspect.
Why the colonial comparison can fit
Calling the book a critique of data colonialism is defensible when its individual stories are connected to structural features rather than to “AI” as a vague synonym for harm.
- Resource appropriation: behavior, speech, images, knowledge and work become inputs for systems.
- Asymmetrical ownership: local people generate data or labor while distant firms control models, cloud infrastructure and commercial returns.
- Knowledge hierarchy: externally designed categories determine what counts as a correct identity, risk or diagnosis.
- Dependency: institutions rely on foreign platforms, datasets, standards and technical services they cannot easily replace.
- Limited consent: participation may be formally voluntary but practically unavoidable when work, healthcare or public services depend on the system.
- Unequal visibility: people are intensely measured while the institutions extracting and classifying them remain opaque.
A 2026 review of postcolonial AI ethics identifies imported infrastructure, proprietary models, external standards and unequal knowledge authority as recurring mechanisms of dependency. The review is available from Springer.
Rank #3
This does not mean that contemporary data extraction is identical to conquest, land seizure or colonial racial rule. One critique warns that “data colonialism” can become too metaphorical and flatten important historical differences. That critique is published by Oxford Academic. The useful claim is narrower: many AI arrangements extend or depend on extractive relationships that resemble older colonial patterns of unequal ownership and authority.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Benefits do not cancel the power question
Jobs and exploitation
Annotation and moderation work can be meaningful and can open paths into formal employment. Its value to workers does not settle who controls the wider value chain or whether wages, safety and voice are adequate.
Access and dependency
An AI diagnostic tool may bring medical assistance to a remote community. The harder governance question is whether that community can inspect the system, adapt its categories, maintain the infrastructure and withdraw without losing essential care.
Accuracy and legitimacy
A model can be statistically useful and still be illegitimate if affected people cannot understand its role, challenge its output or participate in deciding where it is used.
Global reach and uneven burden
Data colonialism is not confined to the Global South. People in wealthy countries are also profiled and managed. But outsourced labor, weaker protections, imported infrastructure and unequal bargaining power can intensify the burden elsewhere.
Rank #4
What the book establishes—and what it does not
Code Dependent offers selected human portraits, not a statistically representative survey of every AI system. Its evidence can show recurring forms of extraction, opacity and reduced agency; it cannot by itself establish the ownership terms, wages, contracts or governance arrangements of every company or deployment.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Nor is there a basis for saying that Murgia formally announces “AI is data colonialism.” The more accurate formulation is interpretive: her reporting can be read through the data-colonialism framework. The book documents situations in which automated systems redistribute power toward corporations and institutions while shifting surveillance, insecurity and the cost of error onto workers, patients, children, migrants and minorities.
Publication and editions
Code Dependent: Living in the Shadow of AI was published in UK hardback on 21 March 2024. The U.S. audiobook went on sale on 18 June 2024, and the book was shortlisted for the 2024 Women’s Prize for Nonfiction. Macmillan catalog pages list different pagination for different editions—304 pages on some listings and 320 on others—so an edition’s ISBN is more reliable than a universal page count. The publisher’s U.S. pages cover the hardcover, digital and audiobook records: hardcover catalog, digital edition and audiobook listing.
The governance question Murgia leaves readers with
The book’s importance is not a claim that technology must be rejected. It is a demand to ask who decides what data is collected, which categories are imposed, where systems are deployed, who receives the benefits and how affected people can refuse or appeal. On that question, Murgia’s human-scale reporting makes a persuasive case that AI’s politics begin long before a model produces an answer: they begin with the extraction, labor and institutional dependencies that make the answer possible.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
