In mid-2024, a cybersecurity partner’s mistake led to a global Microsoft Windows outage. It affected 8.5 million Windows devices. Although this represents less than 1% of all Windows machines, it caused major disruptions. Airports were shut down, major airlines experienced widespread delays, and media outlets were unable to broadcast live content. Emergency services in the U.S. were affected, along with banking services, stock markets, petrol stations, retail stores, hospitals and telecommunications companies. The outage starkly illustrated the extent of the world’s dependence on Big Tech.
The question of (digital) sovereignty — and whether democratic governments can still secure it — comes to the fore. Outside the U.S. and China, Big Tech and especially cloud hegemons have significantly eroded digital sovereignty. Technological dependencies extend beyond individuals and private organizations.
In 2024, the U.K. signed a five-year agreement with Microsoft for its digital transformation. The German government has a similar agreement with AWS and its German partner Bechtle AG. AWS is used around the world, from the Australian Bureau of Statistics, to Argentina’s Provincia Bank, the public bank of the Buenos Aires province. In New Zealand, the Department of Internal Affairs holds an all-of-government agreement with Google Cloud. While most of these public sector deals with cloud hegemons go under the radar, a high-profile example was Israel’s $1.2 billion contract with Google and Amazon to develop a private cloud infrastructure. Called Project Nimbus, the initiative came to light after employees from both companies revealed it was designed to support the Israeli military’s use of AI for surveillance and unlawful data collection targeting Palestinians.
These are just a few examples of a global trend in which governments feel compelled to pursue a digital transformation by outsourcing key control technologies to cloud hegemons. Resulting dependencies complicate efforts to regulate them, as governments rely on cloud hegemons for the very technologies that underpin economic governance.
Among the governments implementing industrial policies to counter the dominance of U.S. (and Chinese) Big Tech, three cases stand out: the European Commission and two approaches from emerging economies. None of these initiatives, however, tackles the core issue of intellectual monopolization, and none has, so far, managed to significantly expand digital sovereignty in relation to the cloud giants and other large tech corporations.
The European Commission’s AI Continent Action Plan, released in April 2025, aims to boost AI development in Europe. On the supply side, the EU is investing in “AI factories” comprising public supercomputers, datasets and other services to support the training of AI models and the development of AI applications. The plan also includes AI gigafactories to be established through public-private partnerships. European startups might train their models on EU supercomputers, but they are still likely to deploy them via Big Tech clouds.
By comparison, Brazil and India have both pursued strategies to develop their own digital technology ecosystems. IndiaAI and the Brazilian AI plan place a much stronger emphasis on the role of the public sector in advancing digital sovereignty. Nonetheless, a shared limitation is that they continue to treat the cloud as just infrastructure as a service, and do not fully decouple from Big Tech.
In the case of IndiaAI, a group of advisors to the government that included employees of Intel and Oracle, recommended the creation of alternative AI infrastructure through an initiative dubbed the FutureLab AI Compute Infrastructure. Echoing the European Commission’s AI gigafactories, this would be developed via a public-private partnership. If adopted, the government will proceed in collaboration with Intel and Oracle, thereby perpetuating structural technological dependencies, which include the reliance of its edtech and healthtech services on Oracle infrastructure.
In Brazil, state-owned telecommunications and data companies have pushed for the development of an intermediate layer to provide computing services to the public sector. Beneath this intermediate layer, U.S. Big Tech companies will continue to supply core black-boxed technologies to the Brazilian public sector. As the trajectory of these initiatives illustrates, it seems to be impossible to decouple from cloud hegemons and other U.S. Big Tech. There has been, though, a seemingly successful story.
The Chinese model #
China is viewed by many as a model worth emulating precisely because it has built its own digital technologies ecosystem including a cloud market dominated exclusively by Chinese players. But even setting aside concerns about massive surveillance and lack of democracy, the Chinese approach is not fundamentally different from the U.S. Big Tech-dominated system when it comes to intellectual monopolies.
In fact, the Chinese model is based on largely imitating U.S. technologies. Chinese tech giants and AI startups have, so far, preferred to avoid risk, positioning themselves as fast followers that build businesses by capitalizing domestically on U.S. advances in AI. Interviewees from Tencent, NetEase, Pony.ai and Megvii described the strength of Chinese tech firms in similar terms: Chinese tech companies do not go from zero to one, but can scale up previously existing discoveries, going from one to 100 at an incredibly fast speed. They adapt the models of U.S. firms to local conditions in order to “cash up,” as one Tencent software engineer put it. An interviewee from Pony.ai noted that Chinese computer scientists trained abroad, and with experience in Silicon Valley, reinforce this culture of imitation when they return to China or work for Chinese firms from abroad.
Fast imitation has been the most viable strategy for Chinese firms when it comes to AI and the cloud, given their access to a huge protected domestic market, rising geopolitical tensions and the entrenched global structure of corporate power. The experience of HuaweiHuaweiHuawei is a Chinese technology company focused on mobile phones and telecommunications, and is seen as a poster child for China’s global tech ambitions.READ MORE serves as a cautionary tale. Although it forged ahead with 5G, geopolitical backlash curtailed its global market expansion. Washington systematically obstructed Huawei’s efforts to roll out 5G across Western economies, allowing competitors time to replicate the technology and restrict the Chinese telecom giant’s market growth.
Overall, geopolitical tensions restrict the potential for any Chinese technological leapfrogging to translate into global market leadership in AI. In this context, the emergence of DeepSeek, a more affordable model, increases uptake not just among end users, but also especially among developers building AI applications. This benefits U.S. cloud hegemons, as DeepSeek exerts downward pressure on other startups, discouraging them from diversifying or establishing independent ecosystems. It also favours those giants because DeepSeek’s models can be rented as a service on their clouds. DeepSeek does not compete with cloud hegemons or disrupt their core business.
But even if China were to leap ahead of the U.S. in technological terms, this would not represent a victory for the majority. It would simply mark a reshuffling of ruling powers. While the turmoil created by such a reshuffling might momentarily distract political and corporate rulers from tightening their control over the rest of the world, the Chinese model ultimately mirrors that of the U.S. too closely to be celebrated or replicated.
Adapted from “The Rulers: Corporate Power in the Age of AI and the Cloud” by Cecilia Rikap published by Verso Books © 2026 by Cecilia Rikap, and reprinted with permission.