State Platform Capitalism

Alongside their dominance in international standards stetting bodies, US platforms also dominate private, de facto standard setting (Giblin and Doctorow, 2022). A core component ofthe platform business model is to anchor de facto standards (such as operating systems or app stores) within technology stacks (Müller et al., 2011). For instance, feature phone internet access across much of the African continent is dominated by Facebook’s Free Basics internet service (banned by India, however, on the grounds of its violation of net neutrality). And Google, through its Google Station program, has embedded itselfin technology stacks throughout the Global South by rolling out free Wi-Fi at over 5000 locations worldwide (Oyedemi, 2021). Similarly, the AI chip designer NVIDIA designed its platform and code library to only interoperate with its GPU semiconductors, making access indispensable for cutting-edge AI applications (including in China, which is striving to develop a fully autonomous AI industry) (Cusumano, 2023). Such approaches accord with Hart and Kim’s (2002, 1) definition of Wintelism as ‘strategies for controlling architectural standards in a horizontally segmented industry’.

China is often accused oftrying to develop a ‘splinternet’ by undermining the mixture of standards and protocols that underpin the open internet (Kenyon, 2021). Huawei’s development of a ‘new $\mathrm { I P } ^ { \prime }$ (a centralized and de-anonymized alternative to TCP/IP, so far rejected at various SDOs) is a prominent example mobilized by critics (Campanella and Haigh, 2023). But the example is often used misleadingly. China is indeed developing an increasingly comprehensive strategy to counter US state-platform power by challenging its chokehold over internet standardization. But rather than trying to fragment this fundamental architecture (Nanni, 2021), China’s strategy is to achieve commercial dominance by becoming a global standards-setting power through technological innovation (Yoo and Mueller, 2024). If successful, its platforms will construct powerful international ecosystems of their own without the need to instantiate ‘walled gardens’ at the network layer.

State Platform Capitalism

55

China has made concerted attempts to establish formal and informal standards of its own (Murphree and Breznitz, 2018). It successfully developed a domestic standard for 3 G mobile communications (TD-SCDMA), which was accepted as one of three viable standards at the International Telecommunications Union [ITU]) (Gao, 2014). Subsequently, China has developed national-level industrial plans to create international standards since the publication of Made in China 2025 strategy by the Ministry of Industry and Information Technology (Malkin, 2020). MiC2025 was followed by the 2018 China Standards 2035 strategy, which in turn generated the 2022 National Standardization Development Action Plan. The latter aims to set global technological standards in a range of high-technology industries, including explicit targets for software, operating systems and AI development (Wu, 2022). The Chinese state has filled positions at SDOs with political appointees rather than corporate delegates (Rühlig, 2023).

China’s concerted efforts have met with some success. China has become amongst the most active members of the International Organization for Standardization (ISO) (Seaman, 2020), while Houlin Zhao served as Secretary-General of the ITU from 2015 to 2022. During the race to develop 5 G technology China successfully achieved ‘technical leadership and the firstmover motivation’ (Kim et al., 2020, 6), establishing the largest portfolio of standard essential patents (SEPs). Huawei has been the top filer of standard essential patents (SEPs) for 5 G to the standards body 3GPP – ahead of both Ericsson and Qualcomm (Ryugen and Hiroyuki, 2020). One recent investigation estimates Chinese firms (especially Huawei and ZTE) will continue to succeed in the filing of next-generation 5.5 G patents, which will act as milestones on the road to 6 G development. However, US efforts to isolate China appear to have weakened its ability to set 5 G standards vis-à-vis rivals like Cisco and Nokia (Kawakami and Miyajima, 2023).

Beyond this drive into de jure global standard setting, China has also sought to establish de facto standards by encouraging adoption of its technologies. Far from its origins as a messaging service, Tencent’s WeChat has become a ‘superapp’ which encompasses many third-party apps (‘mini-apps’). In this way, it has effectively bypassed Android and Apple stores (which suffer restricted access within China) and become a principle interfacing mechanism with the internet for many users (Thun and Sturgeon, 2019). This has enabled Tencent to control a sizeable ecosystem, forcing mini-app developers to conform to its technological standards.

56

Reinventing Capitalism

However, setting international standards has proven more challenging, and requires supporting private internet players in their efforts to embed digital technologies in international technology stacks. As noted in Section 3, ‘bundling’, or full-stack offerings, are a key source of international advantage for China. While its technologies may sometimes be behind the technological frontier, their adoption is nevertheless encouraged either directly by lower costs or by contractual stipulations bundling them with hardware where China is cost-competitive (Seaman, 2020). In this way, the Digital Silk Road serves as a vehicle for setting Chinese technical standards without necessitating their adoption at SDOs. The 20% cost competitiveness in telecommunications hardware offered by Huawei, for instance, incentivizes states to integrate with its networking software ecosystem (Ghiasy and Krishnamurthy, 2020). It also supports adoption of third-party systems, such as China’s BDS-3 satellite navigation system, an alternative to US-operated GPS. CITCC recently constructed Tanzania’s upgraded broadband network to be compatible only with Huawei routers (Agbebi, 2022, 5). Huawei’s HarmonyOS aims to use similar mechanisms to displace Android and Apple, and has so far secured 4% market share in this duopoly (Lan et al., 2022). Inroads into software provision can be more easily made where Chinese internet cabling, data centres, and cloud storage facilities are established (Shen, 2018; Munn, 2020). The rollout of smart city technology, for instance, boosts Chinese software embeddedness across economies and territories more generally (Feldstein, 2020), and allows for the assembly of data from social media and payment apps WeChat and Alipay (Ekman and de Esperanza Picardo, 2020). Similarly, SinoPharm and SinoVac’s vaccine provision under the ‘health silk road’ have bolstered Alibaba’s overseas expansion into third countries’ digital infrastructures, because they are predicated upon cold-chain storage and delivery of vaccines provided for by Alibaba’s cloud-computing system, via the Alibaba Health platform (Habibi and Zhu, 2021).

As de facto standards become adopted, they act as a source of pressure towards aligning ‘global technology standards developed by bodies such as the International Telecommunication Union (ITU) with proprietary technologies used by Chinese suppliers’ (Robert and Paul, 2020). Aware of how standard-setting capacities have cemented the dominance of its platform giants, US state agencies are increasingly collaborating with big tech firms to limit the space for Chinese (public and private) actors to continue to influence standards. Plantin (2021), for example, cites the case of Open RAN. Nominally a set of standards to ensure interoperability between components in a 5 G network, the Open RAN ‘community’ mostly consists of US-sponsored SDOs which exclude Chinese players. These groups (such as the Open RAN Policy

State Platform Capitalism

57

Coalition) enjoy backing from the US Federal Communications Commission to work to devise standards enabling the unbundling of5G hardware and software, in ways which enable US platform giants such as Facebook and Google to collaborate with Huawei rivals Nokia and Ericsson to freeze out Chinese competition.

Similarly, in the AI sphere, Chinese technology firms have become major exporters of AI, as it integrates such services into broader technology and infrastructure deals in areas like health, communications technology, and e-governance. China has successfully captured market share on the African continent by supplying ‘cloud storage capabilities, building out data centers, and then integrating all of those technical systems with AI-capable tools’ (Bouey et al., 2023, 28). Although smaller than giant US rivals Amazon and Microsoft, Alibaba and Tencent have become major global cloud service providers. They operate large networks of global data centres, which are now being integrated with these platforms’ growing AI capabilities such as Alibaba’s Qwen (TechNode, 2023; Dobberstein, 2024). China’s increasingly competitive position in global AI services provision raises the possibility of its firms being able to set de facto standards in the global industry (Drexel and Kelley, 2023). Domestically, China’s regulators have already begun to develop a suite of technical standards for compliance with its increasingly expansive AI and algorithmic regulations (such as techniques for watermarking AI-generated content) (Sheehan, 2024). If these become embedded in its platforms’ global offerings, they are likely to become embedded (Rühlig, 2023).

But in AI, as in networking equipment, the United States is using extant standards-setting power as a cudgel with which to weaken China’s competitive threat. As Bloomberg recently reported, ‘[s]ome American lawmakers have argued that the United States should play a leading role in the shaping of global AI governance, pointing to China’s advancements in the field as a safety and competition concern’ (Saraiva, 2024). To this end, the United States recently established an Artificial Intelligence Safety Institute (AISI) within its National Institute of Standards and Technology. The AISI launched a state-industry consortium for AI safety in February 2024, packed with over 200 industry representatives including Google, Meta, Amazon, NVIDIA, and others (Shepardson, 2024). The consortium will help the AISI fulfil its strategy, shared in outline form in May 2024, to ‘ensure that AISI’s impact is sustained, domestically and internationally, through the adoption of AI safety norms’ and standards (Artificial Intelligence Safety Institute, 2024). As it collaborates with its domestic platform giants to produce technical standards for the AI industry, the US government is also driving international initiatives designed to cement its status as a leader of AI standards. It has played a leading role in the

58

Reinventing Capitalism

G7 Hiroshima Process for developing international AI standards, which notably excludes China (Habuka and Osa, 2024). At the same time, it recently drove the adoption of the UN’s first resolution on AI governance, which calls for international standards for AI interoperability.

To the extent that the United States intervenes to restrict China’s ability to standardize its technologies through formal bodies and mobilizes connections between US state institutions and platform giants to head off Chinese competition, Chinese firms and state agencies would seem all the more likely to double down on whatever de facto standardization avenues remain open. This implies a potential for greater fragmentation of global standards, and bodies such as the BRI’s Standards Organisation – yet to get off the ground – may be mobilized more actively. This would mean that China establishes a series of de facto standards, while the United States remains able to set de jure standards. If this happens, which set ofstandards will be more widely used, how they will interact with one another and what the impact will be for the internet as a whole will become ever more intensely contested fields.

8 Smart Cities

The mutually reinforcing trends that have propelled SPC – state capitalism and the platformization of ever grater spheres of economy and society – have also shaped contemporary urbanization. UN Habitat, the supranational body’s division that is focused on cities, finalized the New Urban Agenda (NUA) in 2016 after an extensive consultation process with UN member states and civil society stakeholders.⁸ By ratifying the NUA, national governments began to reassert authority over urban policy that they had significantly relinquished during neoliberal reforms. Most importantly, the NUA obliges national governments to establish a national urban policy, whose implementation is delegated to local authorities. In other words, national governments are to make urban policies and city governments are to enact them.

National urban policies increasingly leverage technology in pursuit of sustainable urban development. Indeed, the NUA committed states to the rollout of smart city programmes. Smart cities employ information and communication technologies to gather data about the circulation of people and things within cities. Various types of sensors are typically used to collect information, which is acted upon, sometimes in real time, by automated systems. For example, traffic lights can adjust to alleviate gridlock and clear the way for emergency vehicles. Alternatively, people may monitor incoming data in control centres, and they are alerted to anomalies by smart city systems. They can review security camera footage and dispatch emergency services if necessary.

State Platform Capitalism

59

⁸ The UN General Assembly ratified the New Urban Agenda in December, 2016. The full text of the agenda can be found here: https://habitat3.org/the-new-urban-agenda/.

Smart cities are paradigmatic examples of platforms insofar as they combine software and hardware in order to intermediate and influence interactions. Some are out-of-the box greenfield projects that are conceived as ‘whole system’ smart cities from the outset (Gaffney and Robertson, 2018). But greenfield smart cities are few in number, and most smart city technology is incorporated into the urban environment incrementally (Shelton et al., 2015). For example, a city may experiment with an automated traffic control system, while other systems remain decidedly analogue.

Smart city systems make universal assumptions about how cities function, but these are often at odds with particular characteristics of actual urban environments. This explains why smart city systems have oftentimes performed disappointingly, exposing a jarring disjuncture with the techno-utopian ideals celebrated by their corporate proponents (Irazábal and Jirón, 2021). Corporations tried to address this problem by establishing a series of universal standards for smart cities.⁹ This effort was taken up by the International Standards Organization, which disseminates standards to national standards agencies. It introduced a series of standards surrounding the representation of urban data, quality management processes and urban governance (Schindler and Marvin, 2018). The idea was to render complex social worlds into a myriad of legible relations among people and things, that could be observed, classified, and ultimately influenced. Proponents of smart city standards hoped that this would enhance the interoperability and effectiveness of smart city technology – if cities could be reduced to a series of relations, then smart city systems could be deployed in heterogeneous urban environments.

Efforts to establish smart city standards began in the early 2010s, when states were reasserting their regulatory authority and platform firms were scaling up their operations at breakneck speed. However, geopolitical rivalry between United States and China was yet to burst into the open. Rather than a competitor in the global smart city market or a geopolitical rival, China’s rapidly growing cities were seen as a potentially lucrative market by vendors – with one 2013 stock-take of smart cities asserting that they ‘will become the first new industry of the twenty-first century’ (Townsend, 2013, 30). Indeed, China was then undergoing an urban revolution unprecedented in human history, and while Beijing made large-scale investments in urban infrastructure in response to the financial crisis, municipal governments in China were responsible for meeting growth targets (Hsing, 2010). Many municipal governments eagerly embraced smart city systems, while corporations from the OECD went to great lengths to secure a first-mover advantage in China. For example, South Korea’s Samsung coordinated the construction of a greenfield city – Songdo – which served as a ‘test bed’ for technologies that purportedly presaged the future of ‘the new China’ (Townsend, 2013).

60

Reinventing Capitalism

⁹ ISO 37120, 37101, and 30182.

But selling smart city solutions in China came with a catch – corporations would have to be complicit in the state’s large-scale surveillance efforts. American firm Cisco apparently didn’t mind. Its software was on display in Songdo, and subsequently packaged in the innocuous sounding ‘Peaceful Chongqing’ initiative. The Wall Street Journal reported in 2011:

The system, being built in the city of Chongqing over the next two to three years, is among the largest and most sophisticated video-surveillance projects of its kind in China, and perhaps the world. Dubbed ‘Peaceful Chongqing,’ it is planned to cover a half-million intersections, neighborhoods and parks over nearly 400 square miles, an area more than 25% larger than New York City. (Chao and Clark, 2011)

Cisco supplied the software for Peaceful Chongqing, but like most smart cities, the system involved a number of firms. For example, Chinese firm Hikvision supplied many of the surveillance cameras. In the following decade Hikvision significantly expanded its operations, and it now offers comprehensive security systems that include hardware and software to firms and governments in more than 150 countries (Hikvision, 2024). Hikvision is not alone, as many Chinese firms have participated in roughly 700 smart city initiatives across the country (Yue et al., 2024).

Many ofthese Chinese ‘hidden champions’ have become global competitors of their erstwhile American, South Korean and Japanese partners. This has provoked a fierce backlash, with critics pointing out that Chinese firms sell surveillance technology such as facial recognition systems that can be used by governments to monitor political opponents and activists (Atha et al., 2020). Emblematic of this coverage was a New York Times story that reported on Ecuador’s use of Chinese smart city technology (Mozur et al., 2019). It explained that a vast network of cameras fed images back to 16 monitoring centres and the country’s ‘feared domestic intelligence agency’. The system, known as ECU-911, was designed and installed by a tie-up between Huawei and the state-owned China National Electronics Import & Export Corporation. The story claimed that there were agreements in place to build similar systems in Venezuela, Bolivia, and Angola. Even more alarmist was an assessment of the operations of Chinese smart city firms in Africa, that concluded ‘[t]he consequences for human rights on the African continent are likely to be dire’ (Gravett, 2022).

State Platform Capitalism

61

Scholars have questioned whether smart city systems built by Chinese firms are genuinely more likely to engender authoritarian governance regimes than technology from their competitors based in the OECD. In reality, governments tend to rely on a range of stakeholders to incrementally deliver smart city technologies into built environments. Rather than contract a Chinese firm to build a ‘whole-system’ smart city, most cities integrate a patchwork of interlocking systems that evolve over time that include software and hardware from numerous firms. Various systems (e.g. traffic monitoring, facial recognition systems, smart energy management) are provided by ‘combinations of Chinese, local, and even foreign actors [that] have participated in the initiation, definition of requirements, technological design, implementation, management and operation, and upgrade of the projects’ (Facundo Vila Seoane and Morena Álvarez Velasco, 2024, 155). All of these have the potential to support potentially authoritarian ends, regardless of their state of origin.

Chinese firms have principally captured a significant share of the smart cities market because oftheir competitiveness. Far from engaging in nefarious authoritarian forms of solidarity, Chinese firms are simply responding to demand from clients – some of whom may have authoritarian aspirations (He, 2024; Majerowicz and de Carvalho, 2024). A more serious criticism is Chinese platforms firms is that they can establish monopoly positions that crowd out local tech firms and inhibit learning-by-doing as they become embedded in urban software systems (Hinane El-Kadi, 2024). And in their efforts to rapidly expand, Chinese firms receive significant support from the state. This includes preferential financing and bundling with hard infrastructure provisioning, as well as access to public funding for research and development.

Alibaba’s activity in Kuala Lumpur is illustrative of these dynamics. It initially focused on a series of investments in discrete systems, such as cloud computing and digital logistics management. These were integrated under the Electronic World Trade Platform, whose purpose was to facilitate access to markets and value chains for small and medium-sized enterprises. This has subsequently evolved into a comprehensive smart city initiative called City Brain, that Alibaba first pioneered in Hangzhou. It integrates a series of platforms, facilitating customs clearance and logistics, as well as digital traffic management and ‘the transplantation of Alibaba’s entire range of cloud-based solutions’ (Naughton, 2020a). Importantly, Alibaba’s City Brain embeds Chinese digital platforms within Malaysian infrastructure, and this, more than the potential for abuse by autocrats, is what worries US strategists.

Platform initiatives toward integration of numerous disparate systems into a single citywide smart city supersystem that includes software, hardware and infrastructure, heightens security risk by expanding the ‘attack surface’ open to hostile agents. A report jointly issued by cybersecurity services of the United States, Canada, the UK, Australia, and New Zealand explains:

62

Reinventing Capitalism

This expanded attack surface increases the opportunity for threat actors to exploit a vulnerability for initial access, move laterally across networks, and cause cascading, cross-sector disruptions of infrastructure operations, or otherwise threaten confidentiality, integrity, and availability oforganizational data, systems, and networks. (Cybersecurity and Infrastructure Security Agency, 2023, 4)

In the context of this expanded ‘attack surface’, US-aligned policymakers often argue that the inclusion of Chinese smart city technology offers a ‘back door’ to critical infrastructure systems. Consequently, the US-China Economic and Security Review Commission commissioned research on Chinese smart cities, and the resultant report recommended that the United States ‘should undertake measures to insulate certain types of critical smart infrastructure from supply chain infiltration by Chinese companies, especially core telecommunications infrastructure, electrical power grid systems, and other categories that together make up the foundation of smart city development’ (Atha et al., 2020). Doing so would require the creation of multilateral financing vehicles involving US policy banks and allies like Japan and Korea. Even before this report was published, the US-ASEAN Smart Cities Partnership had been announced,¹⁰ and some of its initiatives are indeed funded by the Development Finance Corporation (DFC).¹¹ The partnership is not limiting its support to countries in ASEAN – for example, it recently supported a small project in Rio de Janeiro.¹² But this is one of many projects in a crowded field – Brazilian cities incorporate smart city technology from firms in the United States, China and many other countries (Majerowicz and de Carvalho, 2024). This mirrors developments elsewhere. Indeed, with few exceptions, most cities will incrementally adopt smart city technology rather than embrace wholesystem solutions. To this end, they will combine US and Chinese technology, and in this context American and Chinese firms will jockey for position. For both the Washington and Beijing, smart cities will remain a focal point of competition as they compete to control the integration of digital systems with infrastructure networks.

State Platform Capitalism

63

¹⁰ www.usascp.org/.

11 www.dfc.gov/media/press-releases/dfc-approves-1-billion-investments-global-development.

12 www.dfc.gov/investment-story/reducing-emissions-introducing-smart-city-infrastructurebrazil.

9 Conclusion

We have made the case that the economic crises of 2008 and 2020 gave rise to two powerful new socio-economic phenomena: state capitalism and platform capitalism. The new state capitalism, forged in response to disillusionment with neoliberalism in the context ofthe global economic crisis, has seen states bolster or revive older mechanisms of intervention like state-owned enterprises and subsidies. Meanwhile, they have experimented with newer techniques such as taking golden shares in private firms and pursuing ‘market-conforming’ industrial policies. These have worked well for individual states in the short run, offering the possibility of a rupture with neoliberal orthodoxy. But their cumulative and long-run impacts appear more concerning – not least because their beggar-thy-neighbour economics and spiralling definitions of ‘national security’ generate mounting and increasingly disruptive geopolitical tensions.

At the same time, the business sector has been transformed by the success of a small number of giant firms pursuing platform capitalist business strategies. Rather than trying to compete in particular product lines, these firms intermediate transactions at scale by enclosing the circulation of capital and exchanges within their proprietary digital ecosystems. Within their platforms, they design and control the protocols governing interactions among users. As such, they have become increasingly infrastructure-like, as businesses and societies are increasingly dependent upon their privately owned digital ecosystems. Platforms, moreover, are highly spatially concentrated – with the large majority based in a handful of cities in either the United States or China.

Each trend has separately received considerable scholarly attention and analysis, but few have stopped to ask exactly how the resurgence of states as powerful and interventionist economic actors relates to the rise of giant, technologically sophisticated digital platform firms whose operations span national boundaries and concentrate power in private hands. Although states and platform firms’ interests do sometimes diverge, our core contention is that their relationships are increasingly close and their interests symbiotic. While platform firms carefully cultivate a ‘Californian ideology’ of opposition to state power, in fact, they have – in the United States and China alike – sought and received extensive state protection and support, especially in their international operations. States have provided steady and reliable custom for platform giants, especially for military and national security purposes, while offering considerable support to their overseas operations. Meanwhile, platforms have embraced connections with states eager to project power internationally – an aim for which platforms are the ideal vehicle, due to their vast scale, governance capacities, and their emerging role as owners and operators of critical digital infrastructure.

64

Reinventing Capitalism