On Google Maps in New Delhi, Jammu and Kashmir sits firmly inside India. Open the same app from Pakistan, and the Line of Control loosens into a dotted line, its status suddenly up for debate. With a single click, a border along a mountain range seems to shift, moving with the eye that looks at it.

This digital malleability marks the most unsettling epoch of territorial conquests yet. Although maps have long been instruments of power, Big Tech exacerbates their capacity to recalibrate people’s terms of belonging.

The first form of cartographic violence arrives with early political maps. As wandering tribes gave way to structured empires, central powers needed to make land “legible.” By employing maps, states sought “to arrange the population in ways that simplified the classic state functions of taxation, conscription, and prevention of rebellion,” writes James C. Scott in Seeing Like a State.

During the Age of Discovery, administrators treated maps as objective, scientific records and used them to quantify land. But that objectivity was fiction. With coastlines traced in mathematical detail while vast interiors were left blank, early modern maps effectively erased the presence of indigenous populations, enabling empires to treat these spaces as terra nullius, land rendered empty and therefore ripe for the taking.

After the Treaty of Tordesillas divided the New World, cartographers became masters of geographical manipulation. As geographer J.B. Harley argued, in the hands of European empires, maps functioned "as much as guns and warships... [as] weapons of imperialism." Scott explains that European empires' obsession with straight lines was rooted in their pursuit of ease in "administer[ing]" and "polic[ing]" their "far-flung, polyglot empire[s]," a logic simple enough to be managed from a desk thousands of miles away. Following the collapse of the Ottoman and Austro-Hungarian empires, maps and straight lines divided peoples and cultures as decisively as they did territory. The 1916 Sykes-Picot Agreement, in which Western powers secretly carved up former Ottoman lands with no regard for the people who lived there, set the stage for conflicts that still haunt the Middle East more than a century later.

While states wield competing maps as tools of nationalist pride and geopolitical posturing, for people living along these contested peripheries, cartography dictates access to basic rights, resources, and survival. The map, for them, is a form of daily, lived violence.

For most of recorded history, the power to draw the map was inseparable from the power to govern. In the twenty-first century, Big Tech has acquired this power.

The commercial possibilities of freely accessible, precise location data became evident when, in 2000, the U.S. government turned off Selective Availability, a feature that had deliberately degraded GPS signals to limit civilian use. Within five years, Google had acquired Keyhole Inc., a satellite imaging company originally funded by the CIA's venture arm, turned its technology into Google Earth, and used it to inform Google Maps. Today, companies like Planet Labs, which images the Earth daily, operate alongside firms such as Maxar and Satellogic, deploying large constellations of satellites.

In law, a corporation is treated as a person, able to own property, sign contracts, and sue or be sued in its own name. These entities operate across the jurisdictions of almost all nation states, command revenues that dwarf the GDPs of some regions, and employ more people than many national armies. And like sovereigns, they have learned to use the threat of withdrawal as leverage against the very states that host them.

In 2021, when Australia passed legislation requiring platforms to pay news publishers for content, Facebook responded by blocking all news from its platform for Australian users. With local news outlets suddenly cut off from their largest source of traffic, the government was forced back to the negotiating table. When Google paused its commercial operations in Russia following the invasion of Ukraine in 2022, Russian authorities seized its local bank accounts, forcing the subsidiary into bankruptcy. The country suffered economic losses, and an estimated 100,000 Russian tech workers fled the country by the end of 2022, pushing the Kremlin into a costly pivot toward homegrown technology substitutes. In both cases, states found themselves negotiating with corporations as near-equals.

By accommodating these companies rather than confronting them, states have transferred sovereignty to entities whose only accountability is to the market. The nation-state, for all its faults, can be challenged in court, voted out, or held to account by its own citizens. The companies they work with answer only to their shareholders. And nowhere is that unaccountable power more visible than in the maps themselves.

For these corporations, the map is a product like any other, and its borders are determined by whoever represents the largest market. When Google depicts Kashmir as fully part of India to users within India, and as disputed territory to users within Pakistan, it purchases continued access to advertising markets in both countries while reinforcing each government's preferred narrative. When it depicts Crimea as Russian to users within Russia, it lends the veneer of cartographic authority to state propaganda. These platforms are not picking sides so much as they are coldly ensuring uninterrupted market access.

The erasure of Palestine makes the consequences unmistakable. Where millions of people live, Google Maps and Apple Maps show nothing, only blank spaces and vaguely labelled voids. As scholar Zena Agha documents, GPS platforms "continue to erode the presence of Palestine… [by] simply leaving territories inhabited by Palestinians blank." Google and Amazon's Project Nimbus, a $1.2 billion cloud computing contract with the Israeli government, reveals what is really at stake. Corporations that depend on government contracts of this scale cannot afford to contradict the territorial narratives of the states they serve. Waze, a navigation app owned by Google, steers Israeli drivers away from Palestinian neighbourhoods, designating them as “dangerous driving areas.” While one driver gets a seamless commute, a Palestinian driver on the same road watches a short trip stretch into hours of forced detours. The routing system isolates Palestinian towns, starving local storefronts of passing trade and sealing them off from the local economy.

Such market-oriented digital segregation is not confined to occupied territories, and often plays out in ostensibly undisputed urban contexts too. In Chicago and New York City, Amazon's same-day delivery has been shown to bypass Black neighbourhoods, even when they are encircled by zip codes that receive the service. Ride-hailing platforms like Uber and Lyft compound this, charging surge prices or simply refusing to serve lower-income areas, while sharing ride data with U.S. law enforcement, aiding state surveillance without explicit transparency. Together, these platforms reduce entire neighbourhoods to logistical afterthoughts, starving them of services that surrounding, often whiter and wealthier, zip codes take for granted.

The same algorithmic geography plays out on the screens of Swiggy, Zepto, and Blinkit delivery workers in India. These apps use geo-fencing to confine riders to delivery zones and algorithmic management to monitor their every movement, calculating delivery windows based on the shortest possible path while ignoring traffic, narrow lanes, or weather. Some workers have been flagged for stopping to rest, for taking bathroom breaks, and for any pause that falls outside the algorithm's idea of an efficient route.

The most consequential version of this logic is in predictive policing, where the map does not just deny services to certain neighbourhoods but actively marks them as criminal. Heat maps of predicted crime, sold to police departments and built entirely on historical arrest data, spatialise decades of racist over-policing, flagging marginalised neighbourhoods as inherently high-risk. India is no exception. The Delhi Police launched the Crime Mapping Analytics and Predictive System, or CMAPS, aggregating data from emergency helplines and historical First Information Reports to generate heat maps of predicted crime. Since Indian policing has historically targeted Dalit and Muslim-majority neighbourhoods as inherently suspect, the data CMAPS learns from is already saturated with structural prejudice. As Vidushi Marda and Shivangi Narayan argue in “Data in New Delhi's Predictive Policing System,” this "leads to over-policing areas inhabited by individuals from vulnerable groups," creating a cycle of confirmation bias. It flags these neighbourhoods as hotspots, bringing heavier police presence, producing more arrests, and ultimately generating more data that confirms the original prediction.

In the end, the map precedes reality. For centuries, empires and governments laboured to manufacture this reality, convincing people that a drawn line was, in fact, their jurisdiction. Big Tech has made this process frictionless, and its reach far greater. By outsourcing cartography to Big Tech, states have allowed corporate interests and historical biases to be baked into the infrastructure of daily life, producing a world defined by algorithmic redlining. We exist, now, only as far as the map allows.