articles
issues
support
about
Historicising Enshittification Lineages of the bad internet
Arvind L. U. sep 2026 · essay
Photo by Joi Ito, taken at ICANN 2007. CC BY 2.0.

Cory Doctorow’s enshittification has become part of the zeitgeist.1 Doctorow, laying out the stakes in his eponymous book and a plethora of articles, sketches out a terrifying picture of how capitalism is using the internet to exploit workers, petty producers, and consumers “every instant” and “everywhere”.2 He shows how internet platforms today surveil their workers round the clock, algorithmically intensifying their exploitation without any middle-management — exemplified by Amazon’s exploitation of its truck drivers and warehouse workers3, or Uber’s use of opaque algorithms and psychological tricks.4 The same platforms also extensively surveil their own users, to target a bouquet of advertisements, tailored to the intimate individualised user profiles. Through the burgeoning internet of Things, enshittification has also made the ownership of everyday objects a relic of the past: cars, tractors, beds, refrigerators, washing machines, cradles, and a host of other such objects all increasingly connect to the internet, giving the manufacturer the power to disable certain features or even to brick said devices remotely, should they decide that the customer has breached their terms and conditions. Examples abound, such as Amazon bricking jailbroken Kindles or John Deere remotely disabling stolen tractors.5

Programmatically, it is easy to see the appeal of enshittification as an analytical framework. Since enshittification is coming for absolutely everything and since practically everyone is a potential victim, Doctorow argues, practically everyone has a stake in reversing it.6 Consequently, the potential anti-enshittification coalition could be massive, and virtually unstoppable. Doctorow diagnoses the proximate cause of enshittification as being the mutation of a competitive market capitalism into a capitalism governed by monopolies and monopsonies.7 And his cure broadly involves setting up sound regulatory guardrails to revive and foster healthy competition among corporations: helping make the entire framework rather attractive, even to establishment institutions like the Norwegian Consumer Council.8

While Doctorow’s work is rather valuable for readers who are being introduced to the existence of internet-enabled exploitation, his historicisation of enshittification is rather unconvincing. First, the “good internet” that he describes, restricted as it was to populations within the global North, took shape in the 80s and the 90s. But this was precisely the early decades of the neoliberal period: accepting this periodisation forces us to view the internet as having emerged, somehow entirely unaffected by the political economy of its age, until the 2000s, when a sudden apparent absence of sound regulations led to its enshittification. And second, the idea that this stage of the internet is an aberration is flawed, and emerges from a blindness to the global South, which has no role to play beyond that of a passive consumer, or a victim of U.S. tech, or as the site of overbearing authoritarian states that strangle the founding values of the internet.

I argue instead that the political economy of the early internet was still nascent, only beginning to take shape at the turn of the millenium, with the evolution of the exploitative internet that we know today. The rapid spread of the emergent internet was also a historically contingent process, moulded by capitalism’s drive to transform the entire world into a global factory, and emerging as a central pillar supporting the globalisation of manufacturing and production. And the dis-enshittification of the internet requires that this entire structure — built to maximise profits through the super-exploitation of the global South — be dismantled, birthing an alternative rooted in struggles against imperialism and capitalism.

Our longer history of the internet begins in the 20th century, with the flourishing of national liberation struggles across the global South. These struggles birthed an era of multilateralism, where the nation-state emerged as the “natural” political unit for the organisation of human society; in principle, world affairs would be adjudicated by multilateral organisations, where almost every nation-state had a vote. Some of these nation-states would attempt, however imperfectly, to reduce social inequality through various forms of economic redistribution, while others yet would align with the explicitly anti-capitalist Soviet bloc. All in all, these developments challenged and disrupted the more explicit flow of raw materials and surplus from the global South to North that characterised the formal colonial period.9 As cross-border solidarities between national working classes strengthened, the genuine potential for anti-capitalist uprisings in the Western bloc forced capital to compromise with labour in the metropolis, birthing both Anglo-American Keynesianism and European social democracy.

It was amidst this landscape that the United States emerged as the de facto leader of the Western bloc, as a dominant manufacturing hub and financial hegemon, as the preeminent military superpower, and as the global epicentre of scientific and technological progress. The U.S. was also home to an overwhelming majority of the world’s leading multinational corporations. The upheavals of the 70s — the Volcker shock, the birth of the petrodollar, the weakening of both the Soviet Union and the non-aligned movement, the normalisation of relations between the U.S. and China — all coincided with a strong push for global economic liberalisation on part of the American state. Domestically, labour movements were significantly weakened; abroad, the United States renewed its aggressive advocacy against multilateralism and for the dismantling of trade barriers, creating the conditions for American FDI across the global South.

In the field of communications, this took the form of attacks on government ministries of posts, telegraphs and telephones (PTTs) across the world. Telecommunication services saw privatisation, first across Japan and Western Europe, and then across the rest of the world, often in the teeth of opposition from telecommunications workers. Broadcasting services, too, which until then had functioned largely as state-subsidised public utilities across much of the world, were transformed into advertisement-dependent markets saturated with private firms. Public broadcasters were either privatised or forced to compete within these markets, as state-owned commercial entities. Some prominent examples of this trend were Italy and Chile in the 70s; Argentina, France, Indonesia, and New Zealand in the 80s; and India, Poland, and Nigeria in the 90s.

This, then, was the geopolitical backdrop to the development of the precursors to the internet.10 As U.S. military networks pioneered computer connectivity and spearheaded data networking, data transport settled on using packet switching for transmission, rather than the more familiar circuit switching used by telephone networks.11 Rival standards emerged, such as IBM’s proprietary Systems Network Architecture, or X.25, a technical standard that emerged from the Open Systems Interconnection model. During this period, fundamental questions were being asked and debated: how many data networks would there be? Would they be owned and controlled by PTTs, or by private actors? Would they be adjuncts to existing voice communication networks, or would they be independent? Would different data networks be interoperable? Who would frame rules and regulations for cross-border interoperability?

The enormous size and weight of the U.S. domestic market and the lobbying power of multinational U.S. corporations played a decisive role in ensuring that the nascent cross-border data and information network would be U.S.-controlled and privately owned, with no governance or policy roles for PTTs, or other representatives of national governments. The very same corporations also acted as the biggest vector for the rapid international spread of the privately developed TCP/IP protocol for packet switching: the neutering of PTTs had successfully weakened the rival technical standards that they had championed, particularly X.25.12 TCP/IP’s interoperability mirrored how telephone networks worked in the United States: for decades, multinational corporations had “dedicated” private telephone networks spanning their offices and manufacturing plants in countries across the world; in the U.S., however, these corporations could connect their private lines to one another, and to the broader public network. With TCP/IP, firms could lease private data lines and establish Wide Area Networks, or WANs, so that their offices and plants across the world could exchange messages, files, databases, and software, all in real time. A firm’s private WAN could also connect to the other firms’ private WANs, and to the TCP/IP cross-border public data network. The absence of the national regulations that circuit-switched networks allowed for ensured frictionless corporate operations over TCP/IP. TCP/IP went on to be adopted by universities and scientific institutions; gradually, TCP/IP-based international data networks came to occupy a huge share of data networks across the world, birthing a cross-border data network that came to be known as the internet.

The massive liberalisation drive across the world played a role in determining how the internet’s governance structures evolved. In the 80s, informal and semi-formal groups and agencies — the Internet Architecture Board (1983); the Internet Engineering Task Force (1986); the Corporation for National Research Initiatives (1986); the Internet Assigned Numbers Authority (1988) — were in charge of how the internet was engineered and managed, and how network and webhost addresses were assigned.13 Each of these entities was subject to U.S. jurisdiction. Eventually, in the 1990s, a new California non-profit, subject to the authority of the U.S. Department of Commerce, took over most of these roles. Right from its inception, the Internet Corporation for Assigned Names and Numbers (or ICANN) followed a so-called multistakeholder model, through which technologists, businesses, and civil society organisations could weigh in on issues related to the governance of the internet. National governments on the other hand, despite being formal representatives of their citizens, could play no role in shaping policy for the internet, a global commons of unprecedented reach. Similarly, the internet Society, another American non-profit with no multilateral component within its governance structure, became the parent of the Internet Engineering Task Force: the guardian of the TCP/IP specification.

These developments once again went hand-in-hand with seismic global reconfigurations: the reunification of Germany, the disintegration of the Soviet Union, and China’s entry into the capitalist order. Over the next few decades, global finance rampaged through the periphery. FDI inflows to the global South rose to form a significant chunk of world GDP. In sync, close to a billion landless labourers and proletarianised peasants were absorbed as cheap labour, while even larger numbers joined the reserve army of the unemployed. The global labour force increased from 1.9 billion in 1980 to 3.1 billion in 2007, with a concomitant shift in the location of global industrial manufacturing: by 2010, four-fifths of the world’s industrial workers resided in the global South, particularly in China.14

For the same productivity, workers in the global South were paid far lower than their Northern counterparts.15 The presence of the vast reserve army of labour in the global South allowed for constant downward pressure on wages, which in turn also began to affect Northern wages. In other words, neoliberal imperialism enabled monopoly finance to accelerate the exploitation of workers across the world. This transformation, through which the entire world was turned into a “global factory” in the service of capital, was enabled by a number of technological developments. These included containerised shipping and, relevant to this essay, the rise of the internet. For a global factory to function smoothly, inputs required at various stages of production have to be sourced from disparate locations across the world; dispersed production processes have to be synchronised; final products have to be distributed across global markets. Only an unregulated, frictionless cross-border data network such as the internet could enable the tracking and coordination of all these diverse activities. As a network, the internet, with most of its traffic routed through the U.S., was also the ideal medium for surveillance, as it allowed U.S. intelligence agencies to spy on communications between foreign governments and foreign corporations.16

U.S. tech monopolies have ultimately always been the primary actors coordinating the global factory. Firms like Microsoft, Apple and Google supply the operating systems for the computers and smartphones the world uses; they developed widely-used software services, such as Microsoft Office or a number of messaging services. With the exception of the German firm SAP, the market for Enterprise Resource Planning software suites, and the bevy of related modules used to synchronise decentralised manufacturing operations — Supply Chain Management, Warehouse Management Systems, Manufacturing Execution Systems, Transport Management Systems — are dominated by American firms such as Microsoft, Oracle, Blue Yonder, Rockwell Automation, or Infor. Many of these firms also supply the software that undergirds the functioning of critical infrastructure, such as electrical grids, water supply, and road networks. Physical computing infrastructure, too, is controlled by American tech firms: from computers (Dell, Compaq, Apple) to microprocessors (Intel, AMD, Nvidia) to submarine cable networks (Google, Meta, Amazon, Microsoft).

The underlying structure of the internet thus very much mirrored the structure of neoliberal imperialism, rather than having emerged, glorious and free, before abruptly falling into a monopolistic hellhole. The specific shape of the internet — its governance structure, underlying protocols and their technical implementations, the topology of submarine cable networks — was a direct result of the geopolitical and socioeconomic forces of that era. And thus shaped, the internet was critical for the global spread of neoliberalism, reinforcing the very same socioeconomic structures that birthed it. Specific policy decisions taken at particular moments by particular individuals may have accelerated this, but the timing of these decisions should be read more a consequence of the temporal weakness of domestic and international opposition. The individuals making these decisions don’t really matter: a Reagan or a Thatcher always comes along when the time is ripe. Nor were these decisions limited to the United States, or to the rest of the Western bloc: rather, deregulation and the simultaneous influx of FDI were thrust upon almost every country in the world. Ultimately, this was what enabled monopoly finance to turn the entire world into a super-profitable production facility, for which the internet was constructed, from the ground up, as a critical tool.

Today, dominant U.S. tech firms themselves function as “global factories”, maximising their profits through decentralised operations. Thus, they tend to set up physical infrastructure like data centres and warehouses in regions that offer them favourable conditions: cheap land, subsidised energy and water, relaxed pollution regulations, and so on. In addition, they tend to take advantage of lax labour legislation, allowing them to contract out most of their work while evading all responsibility for worker welfare. This had led to the outsourcing of not just programming and testing, but also poorly compensated work moderating social media, or creating training data for artificial intelligence models.

A majority of the abuses that Doctorow documents as a consequence of enshittification — Amazon’s exploitation of delivery drivers and warehouse workers, Uber’s exploitation of cab drivers, exploitation of chicken farmers by the monopsonies of the poultry industry, hospitals tuning salaries based on nurses’ financial desperation — are all better read as examples of the exploitation of workers by their employers.17 In all fairness, Doctorow himself acknowledges this: “Jeff Bezos neither cherishes his coders’ kidneys”, he writes; “nor is he especially hostile to delivery drivers’ need to pee — he just squeezes any and every worker in any and every way he can.”18 Despite this, the principal cure prescribed for enshittification is antitrust legislation — which has no bearing on workers’ rights, nor on the dynamic between capital and labour, instead limiting itself to the sphere of civil rights. The downstream effects of such legislation, ranging from privacy to adversarial interoperability to the right to repair, are consumer-centric effects that imagine society as a collection of individual, atomised consumers. And while it is undoubtedly a good thing for people to have privacy on the internet, or to be able to access Tiktok from their Facebook account, or to be able repair their iPhone, these are not remedies for the economic policies that lead to their starvation or destitution in the first place.

Such an analysis also tends to situate the European Union as ground zero for a post-American internet, as the most potent force for regulatory actions against tech giants and their anti-competitive behaviour.19 Since imperialism is not part of this framework, the EU itself does not register as a site for the reproduction of the metropolis/periphery dynamic. The brutal austerity and structural adjustment measures imposed upon Greece, Spain, and Portugal by German-dominated European institutions are a classic example of this mechanism in action, as is the role played by the EU as a neocolonial power in its own right. The EU’s inhumane border policies, too, lead to the death of thousands of refugees fleeing wars and societal collapse that EU policies contribute to.20

Simultaneously, while overestimating the role of the EU, enshittification also fails to register other interventions restructuring the internet today. It does not engage at all with the evolution of the Chinese internet, the only iteration truly independent of U.S. hegemony. Nor does it engage with the consequences of the weaponisation of the internet by the U.S. and Israel, actions that have contributed to internet restrictions in Russia and Iran.21 Each of these developments has implications for the functioning of the global factories coordinated over the internet.

In the end, enshittification as a starting point does little to imagine an alternative political economy for the internet. Nor does it question whether internet corporations ought to exist at all. “The problem with the platforms isn’t that they abuse their immense power,” Doctorow writes. “It’s that they have that immense power.” As to why they have this immense power: “A hundred companies are a mob, a rabble. Five companies are a cartel […] The pro-­monopoly policies of the past forty years have produced gigantic companies that find it easy to unite against their regulators, even as the deregulatory policies over the same period have starved regulators of the resources they need to fight back. The inevitable result is regulatory capture.” It is precisely this understanding that informs Doctorow’s analysis of today’s internet as a by-product of a capitalism gone berserk, and the solution being to bring back the competitive market capitalism of a bygone era.

But these competitive markets are simply part of capitalism’s mythology. The tendency to centralise is built into the structure of capitalism. As Baran and Sweezy show in their 1966 book, Monopoly Capital, every industry that reaches a certain level of maturity immediately consolidates around a few players who then effectively act as a monopoly or a monopsony, structures that enable them to accumulate vast surpluses.22 Antitrust actions and other regulatory measures, to the extent that they do emerge, should be viewed as state efforts to maintain the delicate equilibrium between the monopolies that rule various industries. Since capitalism is an intrinsically unstable system, prices and profit margins in certain industries may shoot well past the average rate of profit, necessitating state intervention to restore some form of equilibrium. The spate of regulatory actions across the Western world today are, therefore, efforts meant to bring the prices and profit margins of the tech industry closer to broader societal averages.The regulatory momentum that Doctorow believes is a response to enshittification is simply a shuffling of the chairs on the same deck; a reconfiguration of monopolies that does nothing to eliminate them. This is simply not their purpose under capitalism.

To his credit, Doctorow does advocate for trade unions as part of the remedy to enshittification. Yet here too, he focusses on trade unions in the United States, particularly on those unionising tech workers. Anti-capitalist agitations by trade unions in the global North are indeed necessary and valuable, but they also raise uncomfortable questions. Will these unions also be anti-imperialist? Will workers in the global North also fight against the exploitation of their comrades in the global South? History offers little by way of hope; numerous studies illustrate, for instance, the ties between post-war union leadership in the U.S. and the military-industrial complex, and how said leadership worked with the U.S. government to repress progressive movements across the global South.23 The internet today is enshittified because it is fundamentally woven into the warp and weft of imperialism and capitalism. We cannot build a better internet within such a framework: we must instead begin by asking how to establish egalitarian societies based on social and economic equality, and how to get there from where we are today. It is only then that we can even begin to imagine what kind of internets might be possible or desirable.

Such imaginings might think of the internet as a public utility, rather than as a commodity provided by corporations. PTTs, the remnants of which still exist in a number of countries, could thus become the primary providers of internet services as public utilities. PTTs were, after all, central hubs for numerous experiments on building cross-border data networks. Their workers were at the forefront of the agitations against the privatisation of telecommunications. Much of that institutional know-how and technical expertise can still be resurrected and redeployed. Should such a model see successful globalisation, it would undoubtedly pose an existential threat to present-day internet monopolies, potentially transforming today’s multistakeholder model of internet governance into a multilateral framework.

There is of course the risk that governments will turn these public utilities into panopticons, asserting their power over internet utilities to intensify the surveillance of citizens. But the fight to reclaim popular control over governments must include the fight to transform the internet into a public utility. We cannot avoid that fight by giving way to corporate ownership of the internet: consider how today, state surveillance is often the product of active collusion between governments and internet monopolies. internet activists, in turn, must also recognise that ongoing anti-imperial and anti-capitalist movements are also movements for a better internet, and make their activism a part of those movements. If internet monopoly corporations were forced to pay market rates for land and electricity and water, and if they were compelled to pay fair wages and overtime to workers across geographies, they would simply not remain as profitable as they are today. Many might even go bankrupt. And those that do remain solvent would be unlikely to possess the power that they are able to wield today.

It is when these struggles succeed, and when they lead to the collapse of these firms, that alternative structures for the internet might emerge from popular actions and public imagination. These alternatives may well be a patchwork of regional networks (“splinternets”) that struggle to interoperate without friction; they may not match today’s ideal of 24x7 connectivity. Whatever form it would take, however, solely by virtue of having emerged through revolutionary movements for a just and egalitarian society, this new network would be a genuine people’s internet.

Notes

  1. 1. Tory Shepherd, “‘What many of us feel’: why ‘enshittification’ is Macquarie Dictionary’s word of the year”, The Guardian, 26 November 2024.

  2. 2. Cory Doctorow, Enshittification: Why Everything Suddenly Got Worse and What to Do About It, 2025.

  3. 3. Michael Sainato, “14-hour days and no bathroom breaks: Amazon’s overworked delivery drivers”, The Guardian, 11 March 2021; Annabelle Williams, “5 ways Amazon monitors its employees, from AI cameras to hiring a spy agency”, Business Insider, 6 April 2021.

  4. 4. Noam Scheiber, “How Uber Uses Psychological Tricks to Push Its Drivers’ Buttons”, The New York Times, 2 April 2017; Simon Goodley, “Second study finds Uber used opaque algorithm to dramatically boost profits”, The Guardian, 25 June 2025.

  5. 5. Amazon: Mark King, “Amazon wipes customer’s Kindle and deletes account with no explanation”, The Guardian, 22 October 2012; Tim Cushing, “Amazon Quietly Bricked Jailbroken Kindle Devices Last Year”, Techdirt, 24 March 2015. John Deere: Kyle Wiens, “We Can’t Let John Deere Destroy the Very Idea of Ownership”, Wired, 21 April 2015; Olexsandr Fylyppov and Tim Lister, “Russians plunder $5M farm vehicles from Ukraine – to find they’ve been remotely disabled”, CNN, 1 May 2022.

  6. 6. Cory Doctorow, “Enshittification is coming for absolutely everything”, Financial Times, 8 February 2024.

  7. 7. Ashok Kumar, Monopsony Capitalism: Power and Production in the Twilight of the Sweatshop Age, 2020.

  8. 8. Forbrukerrådet, Breaking Free, March 2026

  9. 9. Samir Amin, Accumulation on a World Scale: A Critique of the Theory of Underdevelopment, 1974; Jason Hickel, The Divide: A Brief Guide to Global Inequality and its Solutions, 2018.

  10. 10. The material here is mostly based on Dan Schiller, Crossed Wires: The Conflicted History of US Telecommunications, From The Post Office To The internet, 2023, ch. 11.

  11. 11. In a circuit switching network, a dedicated telecommunications circuit — a scarce resource — is set up between the data sender and receiver. That path remains blocked for the entire session, and cannot be used by any other device. However, the single path of transmission guarantees the arrival of data in the order in which it was delivered. In a packet switching network, an originating message is broken up into and transmitted as a series of packets along multiple pathways. Packets reroute in case a particular route fails. They may arrive out of order, however, they carry address headers which enable their reassembly in the proper order at their destination. Bandwidth is shared, and each pathway can be used by multiple packets originating from different senders and destined to reach different receivers. Circuit switching networks thus guarantee dedicated bandwidth, low latency and predictable performance, while being expensive and hard to scale. In contrast, packet switching networks may have higher latencies and lower quality of service, while being cheaper and more scalable.

  12. 12. Schiller, Crossed Wires, pp. 569-572; see also: History of Domain Names, “X.25 and public data networks”.

  13. 13. A network address identifies an entire (sub)network of computers, while a webhost address identifies a specific machine or interface connected to that network. The Internet Assigned Numbers Authority, which oversees the entire pool of addresses, allocates large blocks of these addresses to the five Regional Internet Registries (RIRs) — covering Africa, Asia-Pacific, Canada/the U.S., Latin America, and Europe/the Middle East/Central Asia — based on their needs. The RIRs allocate sub-blocks of addresses to Local Internet Registries and Internet Service Providers, who in turn assign specific network/webhost addresses to end-users.

  14. 14. Ajit K. Ghose, Nomaan Maji, and Christoph Ernst, “The global employment challenge”, International Labour Organisation Research & Publications, 1 May 2008. John Bellamy Foster, Robert W. McChesney and R. Jamil Jonna, “The Global Reserve Army of Labor and the New Imperialism”, Monthly Review, 1 November 2011.

  15. 15. Intan Suwandi, R Jamil Jonna and John Bellamy Foster, “Global Commodity Chains and the New Imperialism”, Monthly Review, 5 March 2019; Jason Hickel, Morena Hanberry Lobos and Felix Barbour, “Unequal exchange of labour in the world economy”, Nature Communications, 29 July 2024.

  16. 16. Schiller, Crossed Wires, p. 591. Leaked classified documents have shown that as late as 2010, AT&T operated multiple U.S. facilities where NSA operatives tapped and monitored global traffic transiting through AT&T’s hubs, including traffic that did not originate within AT&T’s own network. Similarly, whistleblower Edward Snowden’s disclosures in June 2013 highlighted the sweeping scope of global internet surveillance by U.S. intelligence agencies, as well as the role played by U.S. Big Tech firms as partners in these surveillance endeavours.

  17. 17. Doctorow, Enshittification, pp. 114-115 (on Uber’s algorithmic twiddling), pp. 120-122 (on Amazon’s exploitation of delivery drivers and warehouse workers), pp. 122-124 (on chicken farmers); Cory Doctorow, “Nurses whose shitty boss is a shitty app”, Pluralistic, 17 December 2024.

  18. 18. Cory Doctorow, “Winning coalitions aren’t always governing coalitions”, Pluralistic, 6 January 2025.

  19. 19. Cory Doctorow, “The Post-American internet”, Pluralistic, 1 January 2026.

  20. 20. Corinna Mullin, “Border Imperialism in the Maghreb: Violence, Exploitation, Accumulation and Resistance”, Transnational Institute, 17 January 2025.

  21. 21. Vera Bergengruen, “How Tech Giants Turned Ukraine Into an AI War Lab”, Time, 8 February 2024; Yuval Abraham, “‘Lavender’: The AI machine directing Israel’s bombing spree in Gaza”, +972 Magazine, 3 April 2024; Ken Klippenstein, “Trump Cheers Lethal Doxxing”, 26 February 2026; Daniel Michaels and Dov Lieber, “How AI Is Turbocharging the War in Iran”, Wall Street Journal, 7 March 2026; see also: Keystone Collective, “The internet Kill Switch Iran Built While the West Was Busy Debating Age Verification”, 24 March 2026.

  22. 22. Paul A. Baran and Paul M. Sweezy, Monopoly Capital: An Essay on the American Economic and Social Order, 1966.

  23. 23. Jeff Schurke, Blue-Collar Empire: The Untold Story of US Labor’s Global Anticommunist Crusade, 2024.