Topic
Anthropology and artificial intelligence
Readings on how artificial intelligence reorganises social life, knowledge and relations.
What is AI anthropology?
AI anthropology studies artificial intelligence as part of social life. It asks how people build, use, interpret and contest these systems, and how institutions turn their outputs into decisions. A model is also a meeting point between categories, labour, infrastructures and ideas about what a person is.
At AIthropology Lab, this perspective brings social and cultural anthropology into conversation with current technological debates. We follow questions of power, knowledge and everyday experience across workplaces, public institutions and the territories that sustain computation. Studying AI in this way also means listening to people whose work or knowledge disappears from accounts of automation.
This guide offers a route through our essays and notes. The readings approach different cases rather than proposing a single theory of AI. Together, they invite comparison between what systems promise, how they operate in particular settings and who can question their consequences.
Three questions to follow
- Who defines the categories? What counts as knowledge, risk or a relevant difference when social life becomes data?
- Who sustains automation? Which forms of labour, care and material extraction make apparently autonomous systems possible?
- Who can demand accountability? How do people contest decisions, claim recognition and change the institutions that deploy AI?
A reading route
AI Does Not Evolve Alone
Start with the debate on progress and classification. This essay locates AI within the social relations that enable its development and questions the idea of autonomous evolution.
When prediction is not explanation
Explore the limits of predictive knowledge. The relationship between models, biology and persons raises the question of what is lost when correlation is presented as explanation.
The gesture as data
A reading on labour and automation, focusing on embodied skills and everyday gestures that become data for training systems.
The cloud has a geography
Follow AI beyond the screen. This essay examines its material conditions and the relationships between infrastructure, energy, territory and extraction.
From Truman to Pax Silica
Connect development anthropology with technological geopolitics. The essay asks how the language of security organises unequal positions in supply chains.
Ayiti: Trouillot, AI and a single finite world
From Haiti and Michel-Rolph Trouillot, shift attention towards the lives and histories left out of promises about the technological future.
All articles on this topic
- Note
Africa at scale
The United Nations backs correcting a centuries-old cartographic distortion. Three weeks later, Accra is debating another form of representation of the continent: the one being encoded into artificial intelligence systems.
- Radar
Land, learning and compute: three scenes of AI in Africa
Cape Town, Egypt and Ghana show how artificial intelligence takes on territory: data centres, mass AI literacy and computational sovereignty.
- Radar
When knowledge starts to act
Museums, Paper2Agent and Anthropic’s arrival in Madrid show how AI is beginning to intervene in the infrastructures that organise knowledge.
- Note
Sixteen days' warning won't put out a wildfire
The 2026 Ecomarcha cycled through the burnt landscapes between León and Xixón. What its chronicles record, and what a predictive model cannot reconstruct on its own.
- Radar
Learning not to delegate
Trade unions, PISA, UNESCO, Latin American universities, Brazil, living heritage and the AI industry itself are shifting the debate towards the capabilities we do not want to outsource.
- Note
Open does not mean shared
China is proposing an open artificial intelligence ecosystem around BRICS to expand the technological capacity of the Global South. The initiative promises to reduce dependencies, although infrastructure, capabilities and forms of governance remain profoundly unevenly distributed.
- Radar
AI has already entered Madrid. Politics is still catching up
An analysis of the place artificial intelligence occupies in Madrid’s politics, administration and public life.
- Note
Who owns a relationship?
From fear of abandonment to the self as a mine. What happens when a company can change who is on the other side
- Radar
When AI needs land
Automatic control, data centres and local media show that artificial intelligence does not live in the cloud. It depends on physical infrastructure, public resources and content over which bargaining power is distributed very unevenly.
- Note
The right not to appear
From the human safari to the algorithmic gaze: when Indigenous isolation becomes content and refusal of contact acquires value.
- Radar
Landscape, truth and a mediated world
Anna Lowenhaupt Tsing, Sherry B. Ortner and Tim Ingold invite us to reverse a familiar question about AI: not only what a technology does, but what relations it needs and what kind of world it helps to produce.
- Note
The self as a mine: Appadurai in Bielefeld
Arjun Appadurai calls the ‘mineable self’ a new commodity made of relationships, stories, profiles and data. A reading of his 2026 work on personhood, affect and artificial intelligence.
- Note
Nokia · The sisu that would not fit in an American pocket
From rubber boots to SMS, from T9 to Asha, and from the collapse of the mobile-phone business to AI networks. A story about Nokia, Finland and the shifting geography of technological power.
- Radar
Palantir and the Digital Twin That Does Not Yet Exist
In 2017, Argentine state agencies explored Palantir tools in connection with the archives of the AMIA case. In 2022, the company registered its trademark in Argentina. Three years later, Próspera co-founder Gabriel Delgado Ayau held a nearly two-hour meeting at the Casa Rosada. In 2026, Thiel Macro disclosed a position worth almost $76 million in Vista Energy, while the government announced a Social Digital Twin that, months later, it acknowledged was not even in the design stage.
- Essay
Why doesn’t Abya Yala have its Accra?
Unthinkable history, extraction and the obligation to repair
- Radar
What we delegate when we delegate to the machine
Three scenes in China, Mexico and Sweden show that delegating to AI no longer means only automating tasks: it also reaches into relationships, the organisation of knowledge and the ways the world is represented.
- Note
Niu Lai and the trace of the hand
A Chinese film made without artificial intelligence went from ridicule to a cultural phenomenon. Between internet 抽象, 手搓 and the desire for 活人感, what happens when imperfection begins to function as a trace of humanity.
- Radar
When a festival becomes a public square. Four days of thinking about the commons at Rototom
Four conversations at Rototom's Social Forum connect mental health, housing, ecological crisis and geopolitics to consider the conditions of the commons.
- Radar
Who retains the power to decide?
Harare, OpenAI and Lahore show that governing AI means deciding who controls data, which institutions retain power, and which forms of knowledge count before automation becomes inevitable.
- Radar
A radar for Federico
Ninety years after his murder, Federico García Lorca has still not been found. This Radar pauses before an absence: the missing body, the mourning that cannot quite end, and a voice that still speaks to us.
- Essay
When prediction is not explanation
From protein language models to the person as data
- Radar
Giving AI a Heart
Languages, culture and society from China, with a counterpoint from Siberia
- Radar
Relationships Come Before the Algorithm
Five frequencies from contemporary anthropology in the United States for thinking about the world we are beginning to automate.
- Radar
What world does AI need? A Colombian reading from Cyberia to invisible labour
Colombia is building sovereign AI models while a critical constellation —Arturo Escobar, Myriam Jimeno, Diana Obregón, Derly Sánchez Vargas and Carlos Andrés Arroyave— helps us ask which forms of knowledge, labour and worlds make that intelligence possible.
- Note
When the cockroaches answer back
In India in 2026, a judicial slur became a political identity. The journey of the “cockroach” opens questions about classification, aspiration, embodiment, satire, power and algorithmic visibility.
- Radar
Who defines, who listens, who learns
Three recent developments in Tajikistan, Kazakhstan, Kyrgyzstan and Uzbekistan show that bringing a language into artificial intelligence also redistributes authority, listening capacity and data labour.
- Radar
Before the model comes the school
Four signals from Chile and Uruguay about the institutions that turn languages, territories and memories into data before an algorithm intervenes.
- Essay
The Agonistic Economy of Digital Capital
Extractive institutions, the technological potlatch and the struggle for pro-worker artificial intelligence
- Radar
Who owns the seas through which artificial intelligence travels?
The expansion of submarine cables shows that artificial intelligence does not inhabit an immaterial cloud. Its infrastructure occupies the seabed, crosses jurisdictions and reorganises relations of ownership, sovereignty and development.
- Essay
When the forest enters the model
Artificial intelligence can improve wildfire detection and response, but it cannot replace the climatic, territorial and social transformation required by an increasingly flammable Europe.
- Radar
Who Governs What Becomes Visible
Three initiatives concerning Indigenous language data, audiovisual heritage, and Kiswahili show that visibility is never mere exposure. It depends on rules governing access, format, quality, and authority, and those rules also shape the artificial intelligence that may be built afterwards.
- Radar
Who Gets Paid When Kiswahili Becomes Data
DataHive Africa is testing a linguistic infrastructure in Tanzania based on paid contributions, community review and consent. The project reveals who speaks, who validates and how value is distributed when everyday language enters artificial intelligence.
- Radar
Technology Comes Later
Three cases show that artificial intelligence does not begin with the model. Before it is trained, institutions and communities have already decided how a language is written, which abilities are recognised as work, and which practices can circulate as heritage.
- Essay
Digital violence is not a design flaw
From synthetic images created in a school to the automatic reuse of photographs on Instagram. An anthropological account of how platforms and institutions distribute harm and the work of repair.
- Radar
Kwe’sx uma kiwe îtxi
A Nasa Yuwe interpreter reveals the potential of natural language processing and raises a decisive question about who governs the data when an Indigenous language enters artificial intelligence.
- Radar
Three Places from Which Artificial Intelligence Is Being Contested
Three cases show how artificial intelligence redistributes labour, knowledge and power across South Asia, Africa and Latin America.
- Essay
From Truman to Pax Silica
How development returned as technological security
- Radar
The radio speaks, the needle remembers and the cloud extracts
Namibia and the Colombian Amazon show that protecting a language or healing a territory also requires decisions about what may become data, what should remain outside the machine and which materials sustain artificial intelligence.
- Radar
What does not fit the mould
Fungi, algorithmic assessment, living heritage and AI ethics in a radar on the costs of making social life governable.
- Note
Cinema ParAIso
Artificial intelligence is reshaping audiovisual production, but the debate is not only about generated images. It affects labour, archives, languages, rights, representation and the power to decide which stories circulate.
- Radar
Kiswahili Language and Global Infrastructures of Translation
Social and cultural anthropology radar on Kiswahili, African language models, AI, memory, heritage and global classification.
- Radar
AI Is Decided at the Border, in Training and Through Contracts
Access to models, scientific training and public procurement show how artificial intelligence is becoming state infrastructure.
- Essay
The Besieged People
How populism turns distrust into an infrastructure of power across platforms, AI, and democratic life.
- Radar
Releasing the Weights Is Not Enough
A model can be downloadable and still remain tied to a border, a cloud or a network. Access to AI depends on weights, permissions and material capacity.
- Radar
The science a platform can read
The documents published by Genesis Mission show that its wager is not merely technical. It redefines what counts as scientific productivity, credit and talent.
- Essay
AI Does Not Evolve Alone
Anthropological approaches to reading artificial intelligence as progress, classification, labour, infrastructure and a struggle over power.
- Radar
Who can claim, who can count, who is left out
Three scenes for considering how artificial intelligence and data reorganise access to justice, the value of science and representation in health.
- Radar
The Machine Sees the Temple. Who Decides What It Means?
ChinaHeritaQA shows that recognising an image is not enough to understand heritage. It also asks who decides which forms of knowledge, categories and narratives count as cultural understanding.
- Note
The Platform and the Category
Palantir, Carissa Véliz and the struggle over who defines what is visible, urgent and disposable
- Radar
A university that made room: social and cultural anthropology at UPV/EHU
A tribute to anthropology at UPV/EHU: gender, bodies, language, art and memory for thinking about artificial intelligence.
- Radar
Ensamblajes Is Not a Waiting Room
Volume 3(1) of Ensamblajes shows why predoctoral research is not a waiting room: music, bodies, the city, myth and death read as situated accounts of social borders.
- Essay
The Third Obligation: Accra and the Debt the West Refuses to Repay
The Accra summit turns resolution A/RES/80/250 into an institutional agenda. This note asks whether recognising historical harm also entails restitution and examines how that question reappears in the political economy of AI.
- Radar
Repair, catalogue, sustain: who keeps custody of memory
From heritage as an object of repair to data catalogued with AI and culture as social infrastructure after crisis: three planes of a single anthropological question about who keeps custody of memory, traced from Africa and Asia.
- Radar
Learning in the public university: five anthropological questions for reading AI from the Complutense
A first-person word of thanks to the teachers of the UCM and a defence of the public university as a space of knowledge, work, equality and debate for thinking about artificial intelligence.
- Essay
War Does Not End Where the Bomb Falls
Ukraine, Iran and an anthropology of peace in the age of algorithmic war.
- Radar
Taking part is not the same as being seen: from territory to data, who reshapes the rules
From Bonn to Busan, Indigenous and community participation enters climate and heritage institutions that are now shaped by data, platforms and AI systems.
- Radar
The frontier is negotiated at Évian
The suspension of Fable 5 and Mythos 5, the G7 summit and the responses from China, Europe and the Global South show that frontier AI is governed through jurisdiction, data, infrastructure and alliances.
- Radar
Translating worlds: climate, heritage and artificial intelligence
Three recent processes show how anthropology helps us read who translates territories, heritages and memories into indicators, archives and AI systems.
- Note
The gesture as data
From Meta to Gurugram, how clicks, skills and bodily techniques are turned into data to train automation.
- Radar
Healthcare AI also has a body
A radar on planetary health, medical AI, embodied suffering and structural inequality.
- Radar
AI does not understand worlds, the forest is not a resource and heritage does not fit inside a file
Three recent scenes show how AI systems, extractive policies and international heritage frameworks translate living cultures into their own languages.
- Radar
Languages, borders and forests. Who has the right to belong
Three scenes connected by a single anthropological question. Who can name, inhabit and care for a territory.
- Essay
Four ways of remapping Abya Yala
A note on Ollantay Itzamná, Davi Kopenawa Yanomami, Silvia Rivera Cusicanqui and Gladys Tzul Tzul, read from Abya Yala and against the colonial map, including the one now learned by artificial intelligence.
- Note
The company that acts alone does not exist
Palantir imagines an armed technological republic, the United States restricts Fable 5 and Mythos 5 as strategic capabilities and Milei proposes non-human corporations. Algorithmic autonomy appears as a political fiction that hides people, data, territory, labour and life.
- Radar
Recognition is not enough: algorithms, territories and disinformation
An anthropological reading of recognition, algorithms, territory and disinformation from Colón, the Brazilian Amazon and Indonesia.
- Radar
China and those who feed AI
Three researchers help us understand Chinese artificial intelligence from the street, the factory and the global infrastructure of data.
- Radar
India under heat: counting, circulating, sowing
An anthropological reading of India through extreme heat, caste census, state data, social circulation and seeds.
- Radar
Three compasses for domesticating AI
Apple, Google and Microsoft did not merely present technology. They staged three distinct ways of making artificial intelligence acceptable.
- Essay
What once destroyed presidents now elects them
From Watergate to AI, an anthropological reading of how visible transgression went from scandal to an infrastructure of power.
- Radar
When anthropology enters the AI design room
From chatbot design at MIT to post-AI pedagogies at St Andrews, anthropology is beginning to intervene not only in how AI is studied, but in how it is designed, taught and inhabited.
- Essay
Fabricated persons
Synthetic figures are not false persons, but fabricated persons. An anthropological reading of how an image comes to act as a person, a moral subject and a bearer of value, through the cases of Lil Miquela, Shudu and Lu do Magalu.
- Radar
The kinship that is sung
Indio Solari, the ricotero mass and a popular form of kinning through music, technology and shared memory.
- Radar
Arturo Escobar and the politics of the pluriverse
From post-development to terricide, an anthropology for thinking worlds in dispute.
- Note
Ayiti: Trouillot, AI and a single finite world
A note on Haiti, international aid and the political economy of artificial intelligence, read through Michel-Rolph Trouillot, colonial memory and human kinship.
- Radar
Marisol de la Cadena and the worlds that do not fit
Indigenous futures, languages at risk and patrimonialised bodies in the June issue of American Anthropologist, the most complete count to date of human remains in British institutions, and Marisol de la Cadena's election to the American Academy of Arts and Sciences: what anthropology does with what the dominant frameworks cannot handle, and what that says about an AI that classifies, translates and normalises.
- Radar
Darío Gil and AI as state infrastructure
The Department of Energy's Genesis Mission situates AI as a state infrastructure of science, energy and security, while Descola, Viveiros de Castro and Veena Das offer anthropological keys for reading what world it produces.
- Essay
It is not enough that it speaks
Linguistics, anthropology and artificial intelligence. Why AI is not an isolated mind but an assemblage of signs, infrastructure and social experience that we would do well to de-fetishise.
- Radar
Spanish anthropology for reading infrastructures, platforms and everyday life
AIthropology Lab radar on Susana Narotzky, moral economy, food delivery, neighbourhoods, networks and commons, from social and cultural anthropology in Spain.
- Radar
Haraway and the case for a slow AI
AIthropology Lab radar on slow AI, funding for the discipline, restitution, more-than-human ecologies and ethnographic repair.
- Note
Not too late again: AI, data and public ethics
The opening in Valencia of the UN's AI Governance Lab, read through anthropology: data, consent, algorithmic harm and public ethics.
- Radar
AI is no longer a tool. It is a social infrastructure
AIthropology Lab radar on regulation, work, education, cultural heritage, the anthropology of AI and material infrastructure.
- Radar
Praxis, care and essential work
AIthropology Lab radar on public anthropology, global health, care work, multimodal pedagogy and kinship.
- Note
Leo XIV, artificial intelligence and the new human question
An anthropological reading of Magnifica Humanitas, the first encyclical of Leo XIV on artificial intelligence, human dignity and technological power.
- Note
A revolution (in this too): weather forecasting
A note on Aardvark Weather and the impact of artificial intelligence on weather forecasting.
- Note
Immortality and the present
A critical reflection on the technological promise of immortality set against the inequalities and forms of violence of the present.
- Essay
Artificial intelligence and the exacerbation of desire: towards an ethical reflection
A critical reflection on AI-generated pornography, objectification, desire, consumption and the representation of sexuality.
- Note
AI and archaeology: revolutionary discoveries in Nazca
A note on the use of deep learning to identify new geoglyphs in the Nazca region.