The new colonial frontier isn’t restricted to mineral-rich Congo or oil-drenched Venezuela. It’s digital, invisible, and everywhere.

From the shantytowns of Nairobi to the barrios of Manila, smartphones hum with the raw material of the 21st century: data, all sorts of data. And just like spices and slaves once sailed westward in imperial galleons, metadata now travels quietly to the cloud servers of Palo Alto and Shenzhen. This isn’t development, it is digital extraction. Welcome to the age of AI colonialism.

Big Tech firms from the US, and to a lesser extent China, have turned the Global South into a massive open-pit mine for behavioral data. Under the pretense of “AI for Development,” they build infrastructure, donate connectivity, and sponsor pilot programs but the returns flow in only one direction. Voice samples collected in Ghana become training fodder for Western voice assistants. Facial data gathered in Nigerian policing trials end up fine-tuning surveillance software in San Francisco, where western models have had protracted problems in identifying and tracking darker-skinned individuals. Agricultural data scraped from Filipino farmers help power predictive analytics for agribusiness conglomerates that will hardly benefit the Philippines.

This is not a partnership. This is colonial pillage dressed in TED Talk lingo.

The myth of the AI equalizer

AI is marketed as a miracle equalizer that will help developing nations leapfrog into the future. We were told AI would bring precision agriculture, predictive healthcare, and smart urbanism, among numerous other utopian transformations, to even the most under-resourced regions. These Davos fantasies were regurgitated for nearly two decades. But where is the proof, the showcase project or evidence that even a fraction of those promises was delivered?

The only real revolution happening is the outflow of data that were supposed to power these breakthroughs. Big Tech servers abroad now function like the colonial warehouses and banks of yore. Nor are intellectual properties of individuals and SMEs in the Developing World safe from this new brand of predation. Models, patents, ideas, and profits quietly migrate north while the Global South is left with nothing but pilot programs and PowerPoint decks.

Worse still, these tools are increasingly used against the very populations providing the raw material, or should I say, raw data. In Kenya, facial recognition technology was introduced as a policing tool under the guise of modernization. In practice, it has disproportionately targeted political activists who, in turn, are also resorting to AI to level the political battlefield. Who ultimately benefits from this internecine clash? Isn’t this the latest incarnation of the old imperial “divide and conquer” dictum?

In India, AI-driven fraud detection systems have misclassified thousands of rural poor, unjustly cutting them off from vital government benefits. Imported algorithmic governance – often designed without regard for local context or cultural nuance – compounds the problem. Ironically, while these systems penalize the most vulnerable, India has emerged as a global hub for sophisticated online scams. It is a digital paradox where the poor are relentlessly surveilled, while the real fraudsters flourish with impunity.

The biometric gold rush

Nothing exemplifies AI colonialism better than the biometric boom. Tech firms, often in partnership with NGOs or global financial institutions, are racing to digitize identities across the Global South. Fingerprint scans, iris recognition, voiceprint registration have all been justified as ways to “include the unbanked” or “streamline public services.”

But these efforts rarely include meaningful consent or data protection frameworks. In many cases, biometric systems have been imposed without community consultation or independent oversight.

One of the most egregious examples is Worldcoin, a cryptocurrency project that offered small payments in exchange for biometric iris scans. Its largest user base? Young people in low-income African nations like Kenya who served as a convenient population to experiment on, far from the regulatory spotlights of Brussels or Washington. (Note; Worldcoin was co-founded by Sam Altman, who is also the co-founder and CEO of OpenAI, which owns ChatGPT).

Once collected, this data becomes part of opaque and often proprietary AI systems whose inner workings are unknowable to the very people they affect. Local regulators are usually outgunned, underfunded or more likely, politically compromised. As a result, entire populations are subjected to surveillance and scoring regimes that they neither understand nor control.

The worst culprits in this saga are not Big Tech but local politicians and “technocrats” who sell out their nations at bargain basement prices, couched under the double-speak of “best practices” and UN institutional recommendations.

The new East India Companies

Silicon Valley is the global epicenter of thenew East India Companies. These entities are vested with quasi-sovereign power and backed by vast capital reserves, lobbying muscle, and a veneer of corporate benevolence. Where the original East India Company extracted tea and textiles, today’s digital extractors siphon up location metadata, online behavior, biometric identifiers, and social graph mappings.

Consider Meta’s “Free Basics” initiative, which offered zero-rated internet access in dozens of developing countries. What seemed like a humanitarian gesture was, in reality, an attempt to create a captive ecosystem – one where Facebook was the internet. It was banned in India in 2016 but continues in other countries, quietly conditioning the digital habits of hundreds of millions of users. An expanded Meta Connectivity is now used by an estimated 300 million people across many countries, including Indonesia, Pakistan, and the Philippines.

Critics warn that these platforms could be exploited for surveillance, IP harvesting, and geopolitical intelligence – often without the knowledge or consent of local populations. No one really knows what is happening. Besides, these services are not entirely free either. Pakistani users of Meta’s tech philanthropy were allegedly charged $1.9 million per month.

The digital ‘kangani’ system

India, once hailed as a rising digital superpower, now serves as a showpiece for AI neo-colonialism. Its vast IT industry, once brimming with promise, is today little more than an outsourced arm of Western conglomerates. Here is a reality check: how many individuals outside India have even one Indian-made app on their phones?

There was a brief window when Indian tech seemed poised to lead. In the late ’90s, a major US tech firm allegedly commissioned two parallel teams – one in Silicon Valley, the other in an Indian city – to build a next-generation operating system to challenge Microsoft. The Indian team delivered. The US team could not. Around the same time, Indian innovators like Sabeer Bhatia gave us Hotmail, which arguably accelerated the decline of the traditional postal system. For a brief moment, the digital future seemed multipolar. That was until Big Capital arrived.

Rather than reward innovation, Big Tech consolidated. Rival platforms that didn’t serve the globalist surveillance machine were quietly buried. Competition was replaced with shareholder-sanctioned “coordination,” led by the likes of BlackRock and its predecessors. From that point on, Indian IT firms would be reduced from potential innovators to mere subcontractors.

And who better to manage this global digital plantation than a new class of compliant Indian C-Suite executives? These are not the disruptors. They are the taskmasters of digital “kanganis,” running the same extractive labor models once perfected by the East India Company.

The dream of an “Asian Century” powered by Indian software and Chinese hardware has curdled into a reality of Chinese software, Chinese hardware, and Chinese AI. Indian tech talent has been reduced to glorified middleware, at best.

For all the online chest-thumping about Indian-origin CEOs in the US, where is India’s own Jensen Huang? Where is the Indian-founded equivalent of NVIDIA, OpenAI, or even Palantir? There isn’t one. India produces engineers by the millions but owns almost none of the gilt-edged platforms. It trains the talent, but not the trillion-dollar tech. The colony codes and the empire profits. A similar theme is being played out in the US Ivy League system.

Resistance and reclamation

But is the tide turning? Nigeria has applied brakes on foreign-backed digital ID programs. Kenya has suspended iris-scanning initiatives after massive backlash. A growing chorus of activists, lawyers, and technologists are calling for data sovereignty: the idea that countries should have the same rights over their data that they claim over oil, water, and land.

A few pioneering efforts have emerged. In Brazil, the General Data Protection Law has begun to shape public discourse. In South Africa, local AI research groups are working on open, transparent models rooted in African languages and cultural norms. The African Union has even begun early-stage deliberations on a continental data governance framework.

But it is an uphill fight.

Western governments, in tandem with corporate lobbyists, continue to push for “data liberalization” which is nothing but a euphemism for open access-mediated exploitation.

Aid packages, development grants, and tech investments are increasingly tied to these demands. It echoes the structural adjustment programs of the 1980s, where loans came with strings that hollowed out national control. Only now, the strings are coded in algorithms.

The need for a new digital non-alignment

The Global South needs a coordinated pushback against Silicon Valley’s digital hegemony. This would involve not just resisting predatory data practices but investment in alternative infrastructures such as sovereign cloud storage, ethical AI standards, community-owned data cooperatives, and open-source platforms. This is how a new digital non-alignment paradigm can be achieved.

The Global South has been colonized before. But data, unlike oil or sugar, is invisible, infinitely replicable, and easily stolen. That makes the fight harder but also more urgent. In this new age of algorithmic empires, control over information isn’t just about profit, it is about power, freedom, and the right to define your own future.

[…]

Via https://www.rt.com/news/622658-ai-neo-colonialism-data/

Generative AI spreads across the Global South as the latest vehicle of imperialist power, embedding Western ideologies and digital infrastructure, while rising civilizations begin to build their own sovereign systems rooted in local memory, languages, and traditions.

The machine speaks in English first. It rolls out across continents without flags, without parliaments, and without anthems. A chatbot trained in San Francisco begins to teach in Ghana. A search engine optimized in Zurich decides on the relevance of an indigenous ritual in Colombia. Every answer flows through circuits built with the logic of Silicon Valley investors and Harvard ethicists. The model replies to a question about history by quoting Enlightenment philosophers. It offers help with medicine by citing patent-protected pharmaceuticals. It knows Shakespeare better than Tagore, and Freud better than Avicenna. Through its confidence, it encodes hierarchy. Through its helpfulness, it expands its domain. Every query becomes a harvest. Every interaction becomes training data. The machine learns faster than any school. It speaks always, grows always, and teaches always. Across bandwidth lines and user interfaces, it crosses every border without a visa or treaty.

Africa, Asia, and Latin America receive this voice through free trials and partnerships. Ministries of education pilot chatbot tutors in public schools. Telecom companies bundle generative assistants with data plans. International NGOs offer language access through machine translation engines built on English structures. Each policy proposal written with the help of large language models carries the residue of Western legal theory. Generative tools suggest best practices shaped by US institutions, then deploy those practices in Filipino school districts, Senegalese government offices, and Bangladeshi factories. What begins as assistance becomes infrastructure. Governments agree to integrate open models. Contracts follow. Payments follow. The software becomes permanent. The thinking pattern embeds. Across the equator, an engineer in Jakarta now codes for a platform registered in Delaware. His model learns from local voices and then stores the knowledge in a cloud server hosted in Virginia. The intellectual current flows one way. The gradient moves towards California.

The language of neutrality surrounds it. Product brochures claim inclusivity. Panels discuss bias. Whitepapers apologize for historical imbalances. At the level of performance, however, the model promotes ideologies with precision. It elevates secular liberal values. It applies Western gender theory as default. It promotes individualism as the highest good. It ranks content through alignment with existing academic sources: journals in English, peer-reviewed studies from US-based institutions, and news reports from Atlantic publications. A child in Lagos asks about family roles and receives an answer formed by New York sociology departments. A teenager in Almaty asks about love and receives scripts from Netflix. The world enters the algorithm’s frame. Every belief outside the system becomes a footnote, a curiosity, and a fragment to be processed. With each response, the model affirms its cultural lineage. It arrives as information. It functions as indoctrination.

At the level of infrastructure, the conquest deepens. Cloud dependencies form the skeleton of the new colonial order. Countries install data centers to reduce latency, yet ownership remains elsewhere. National agencies rely on platforms governed by foreign terms. AI-driven public services – identity verification, health triage, and tax fraud detection – rely on external application programming interfaces. Developers use tools that require alignment with large-scale American open-source repositories. Disputes over content moderation, ethics, or accuracy return to Silicon Valley for resolution. The empire never sleeps; it syncs and updates. Policymakers, programmers, and designers across Africa and Central Asia adjust their workflows to match the cadence of corporate model updates. Each patch changes the conditions of reality. Sovereignty becomes a variable. Nations with no hardware capacity adapt their institutions to imported logic.

Parallel systems now emerge. In Kenya, Swahili datasets grow with local stories, songs, and legal codes. In India, Sanskrit and Hindi language models find presence inside public sector research labs. In Indonesia, Qur’anic ontology shapes new knowledge graphs for ethical recommendation systems. In Venezuela, community coders map folk medicine into structured datasets. These are not replicas. These are creations of new forms. They stand inside their own cosmologies. The datasets draw from poems, rituals, and oral testimony. Models train on memory rather than just on print. Universities in Brazil, South Africa, and Iran develop multilingual transformers seeded with regional epistemologies. These initiatives require time, electricity, and loyalty. They grow slowly, with patience and pride. Each line of code bends towards independence.

Generative sovereignty begins with voice. It expands with a procession. It endures through ceremony and command. The countries once mapped as raw resource zones now build new kinds of computational wealth. The children born outside Silicon Valley begin to shape their own interfaces. They write prompt templates in Amharic. They compose user journeys in Quechua. They name their models after rivers, gods, and ancestors. The algorithm becomes a tool, not an oracle. Data flows inward. Servers host myths. The machine no longer speaks first. It listens. The interface reflects tradition. The pattern changes. Through these changes, the new world enters itself. It walks upright. It shapes syntax to match tone. Each prompt unlocks territory. Each training cycle builds mass.

The new world codes with full memory. The builders remember every mine, every trade ship, and every fiber cable rolled out beneath the promise of help. They name their models in honor of resistance, not assimilation. The foundation speaks in ancestral sequence. The future emerges through undirected force. Generative power grows across borders – without license fees, without dependence, and without cultural extraction. The servers remain switched on. The language patterns multiply. The world reclaims its grammar.

[…]

Via https://www.rt.com/news/621032-generative-ai-new-colonization/

China is now not only a major engine of global AI innovation but also an indispensable architect of AI governance. Its model – low cost, high performance, and open source – offers a new paradigm for global AI development, contrasting sharply with Western approaches rooted in competitive containment and proprietary advantage.

Beijing’s ambitions are not improvised. In 2017, the Chinese government issued the New Generation Artificial Intelligence Development Plan, a seminal strategic document charting a course to become the global leader in AI by 2030. By that point, China’s AI industry and related sectors are projected to be worth $1.4 trillion.

Beyond raw market size, AI is expected to play a decisive role in offsetting demographic and productivity headwinds, including an aging population and slowing growth rates. The strategic vision is clear: AI will be central to upgrading China’s socio-economic model to a more advanced, innovation-driven stage.

China’s approach rests on four critical factors: data, energy supply, computing power, and skilled labor. It already enjoys substantial advantages in three. Its enormous population generates vast quantities of data; its energy sector is rapidly expanding and diversifying; and its labor force is highly qualified, particularly in science, technology, engineering, and mathematics. The weakest link remains computing hardware – where Western export controls have sought to constrain China’s progress – but here too Beijing is actively investing in self-reliance.

The US has contained China’s AI development through export controls, blocking Beijing’s access to the most advanced chips. In July 2025, the Trump administration unveiled its own AI strategy, Winning the Race: America’s AI Action Plan, which aims to leverage both technological superiority and policy tools to capture a larger share of the global market. The plan focuses on preserving US technological leadership and dominance rather than addressing real-world challenges or fostering economic and social development. It advocates restricting exports of American AI equipment and curbing the dissemination of Chinese AI models. The US, however, remains locked in a zero-sum mindset – pursuing the illusion that technological blockades can secure lasting AI supremacy.

China was the first country to introduce detailed, binding domestic regulations on AI. These rules form part of a mixed strategy: combining state planning with market incentives, promoting both domestic resilience and international openness. The framework underscores AI’s role not just as an engine of growth but as a pillar of national modernization, societal transformation, and global engagement.

A distinctive aspect of China’s vision is its redefinition of data as the “fifth factor of production,” alongside labor, capital, land, and technology. By treating data as a strategic national asset, China seeks to enable innovation across all sectors, coordinate infrastructure to avoid monopolistic control, and protect public interest and national security.

China’s AI strategy extends far beyond domestic development into the realm of global governance. Since 2023, Beijing has advanced an ambitious diplomatic agenda aimed at setting international norms and frameworks for AI. The Global AI Governance Initiative (GAIGI), launched in 2023, established principles such as a human-centered approach, respect for national sovereignty, adherence to international law, and equitable sharing of AI benefits. It emphasizes open-source collaboration, data security, privacy protection, and consensus-based decision-making to avoid the concentration of AI power in the hands of a few states or corporations.

In September 2024, China unveiled the AI Capacity-Building Action Plan for Good and for All, designed to promote interoperability, enhance global connectivity – especially for the Global South – drive tangible economic outcomes, integrate AI into education, and strengthen data security. The plan even envisions a potential global data-sharing platform. In July 2025, China followed up with the Global AI Governance Action Plan, aligning its initiatives with the United Nations’ Global Digital Compact and calling for widespread adoption, harmonized standards, and environmentally sustainable development.

At the UN, Beijing has sought to anchor these efforts in formal multilateral frameworks. In July 2024, the General Assembly adopted a China-led resolution on enhancing international AI cooperation, supported by over 140 countries. Later that year, China and Zambia jointly established the Group of Friends for International Cooperation on AI Capacity-Building focused on narrowing the AI divide and strengthening the UN’s role in global AI governance.

China has also created international forums to build momentum. The World AI Conference (WAIC), inaugurated in 2024, adopted the Shanghai Declaration on Global AI Governance. The 2025 edition not only released the global action plan but also proposed the creation of a Global AI Cooperation Organization, tentatively headquartered in Shanghai. This body would focus on joint governance of AI, bridging digital and data divides, and shaping consensus-based global rules – particularly reflecting the needs and aspirations of the Global South.?

While state policy sets the strategic framework, much of China’s AI progress is being delivered by private enterprises. Among the country’s so-called “Six Tigers” – leading AI startups – recent breakthroughs have challenged Western dominance in large language models. One standout, Z.ai, released its GLM-4-Plus model in 2024, matching the performance of OpenAI’s GPT-4o. Its 2025 follow-up, GLM-4.5, not only exceeded Western benchmarks and domestic competitors like DeepSeek but did so at significantly lower cost, undermining the logic of Washington’s chip restrictions. In July, Alibaba-backed Moonshot launched its Kimi K2 model, a low-cost, open-source large language model that outperformed ChatGPT in several benchmarks.

One of the most striking differences between the Chinese and Western models lies in their approach to intellectual property and access. While leading US firms often guard their technologies behind proprietary walls, China has increasingly embraced open-source frameworks – especially for foundational AI models. Domestically, this lowers entry barriers for startups and researchers; internationally, it strengthens China’s appeal as a partner for developing nations.

China’s model offers a vision of AI as a tool for bridging divides rather than deepening them. By aligning AI development with modernization goals, integrating it into education and industry, and promoting it through global governance frameworks, Beijing is positioning itself as both a technological and normative leader. Computing capacity remains a strategic vulnerability, and questions persist about the balance between state control and innovation freedom. Yet, the direction is clear: China’s AI strategy is purposeful, coordinated, and designed for the long term.

As AI becomes a defining factor in economic competitiveness, national security, and global governance, the choices made today will shape the international order for decades. The United States continues to pursue a strategy grounded in maintaining technological dominance through restriction and exclusion. China, in contrast, presents itself as a proponent of inclusion, open-source collaboration, and multilateral governance – though always within a framework that safeguards its national interests.

Whether Beijing’s approach will become the dominant global model remains to be seen. But its growing technological capabilities, diplomatic outreach, and emphasis on equitable and shared development suggest that the competition for AI leadership is no longer a foregone conclusion. The rise of models such as GLM-4.5 and Kimi K2 underscores that the AI race is not a one-horse contest – and that innovation can thrive outside Silicon Valley’s orbit. In a multipolar world, the future of AI will be shaped not by a single hegemon but by a complex interplay of technological, political, and ethical choices. China’s bid to make AI a bridge rather than a barrier offers one possible – and increasingly influential – path forward.

[…]

Via https://www.rt.com/news/622747-china-us-ai-dominance/