Logo of Microsoft in London, Britain.
Image: XINHUA
Imagine two students, both sixteen years old, both curious about the world. One lives in Oslo, the other in Nairobi. The student in Oslo uses an AI tutor to work through calculus problems, drafts essays with a writing assistant, and practises her French with a conversational AI that corrects her in real time. The student in Nairobi has a smartphone but unreliable data. The AI tools that exist for her (if they exist at all). were not built with her language in mind, her curriculum in view, or her economic context as a reference point.
That gap, replicated across billions of interactions and hundreds of millions of people, is what the data now confirms with uncomfortable precision. AI adoption in the Global North stands at 24.7% of the working age population, compared to just 14.1% in the Global South, and crucially, the divide is not closing. It is widening. The Global North is adopting AI nearly twice as fast, which means every month that passes without structural intervention, the distance between those two students grows a little larger.
This is the context in which Microsoft's announcement at the India AI Impact Summit in New Delhi this week lands significant, necessary, and still insufficient on its own.
Microsoft says it is on track to invest $50 billion by the end of the decade to help bring artificial intelligence to lower-income countries, structured around five pillars: building physical infrastructure, expanding digital skills, developing multilingual AI capabilities, enabling locally relevant innovation, and measuring adoption to guide future decisions.
The infrastructure commitment is real and already underway. Over the past fiscal year alone, Microsoft invested more than $8 billion in data centre infrastructure serving the Global South, including India, Mexico, Africa, Southeast Asia and the Middle East. The company is also working to extend internet access to 250 million people in underserved communities, with 117 million already reached in Africa, a figure that matters because without reliable connectivity, everything else is theoretical.
The skills dimension is equally concrete. After training 5.6 million people across India in 2025, Microsoft has set a goal to equip 20 million people in India with essential AI skills by 2030, alongside the Elevate for Educators programme aimed at training two million teachers across more than 200,000 schools. These are not small numbers. But viewed against the scale of the challenge, a Global South that encompasses roughly six billion people they are a beginning, not a solution.
The multilingual investment is perhaps the most structurally important pillar of all. Language has always been the invisible wall that divides the connected from the excluded. Most frontier AI models were trained predominantly on English-language data, which means they perform worse in Swahili, Tamil, Hausa, Tagalog and the hundreds of other languages that billions of people actually think, learn and work in. Programmes like LINGUA Africa, committing $5.5 million to AI models for underrepresented African languages, represent the right kind of thinking, though the funding figure is modest relative to the linguistic diversity it aims to serve.
There is something quietly revealing in a number buried in the reporting around this announcement. Microsoft invested roughly $80 billion into data centres last year alone, more than half of which was directed to a single economy: the United States. The $50 billion for the Global South, spread over a decade, covering dozens of countries and billions of people, is a meaningful commitment. But the comparison puts its scale in perspective.
This is not a criticism unique to Microsoft, it is the nature of how capital moves. Investment follows expected returns, and the Global South, for all its potential, carries risks, currency volatility, regulatory complexity, infrastructure gaps that make capital more cautious. What Microsoft is doing, in part, is subsidising access that the market alone would not provide. That is genuinely valuable. But it also illustrates why corporate philanthropy and investment, however generous, cannot substitute for the kind of systemic infrastructure spending that only governments and multilateral institutions can mobilise at the necessary scale.
There is also a geopolitical dimension that the New Delhi summit did not fully confront. DeepSeek's strongest AI adoption has emerged across China, Russia, Iran, Cuba and Belarus, but perhaps even more notable is its surging popularity across Africa, where it is aided by strategic promotion and distribution through partnerships with telecom services. The AI race is not simply a development story. It is a contest for influence, and the Global South is its primary arena. Chinese technology companies have moved aggressively into markets that Western platforms have been slow to reach, free of cost, free of the credit card requirements that locked millions of potential users out of early AI adoption, and strategically distributed through existing infrastructure relationships.
The question for the Global South is not simply whether to get access to AI. It is whose AI, built on whose data, governed by whose rules, and serving whose interests.
The honest answer is that closing the AI divide requires things that no single company can deliver. It requires reliable electricity, still unavailable to hundreds of millions of people across sub-Saharan Africa. It requires broadband infrastructure, which in many rural communities remains prohibitively expensive even where it technically exists. It requires regulatory environments that encourage local AI development without creating barriers to competition. And it requires training data that reflects the lived realities of communities in the Global South rather than simply translating the assumptions of communities elsewhere.
India's positioning at the centre of this moment is not accidental. The Indian developer community is the second largest on GitHub at 24 million, growing at more than 26 percent annually since 2020, with a recent surge of over 36 percent in Q4 2025. India is not simply a recipient of AI investment, it is a generator of the talent, the open-source contributions, and the innovation that the global AI ecosystem runs on. Its ambition to lead the Global South's AI agenda is credible, and the AI Impact Summit, hosted by Prime Minister Modi and drawing the CEOs of OpenAI, Anthropic and Google alongside Microsoft's leadership, signals that the world's major technology players recognise this.
But leadership means more than hosting impressive summits. It means insisting that the AI models deployed across Global South markets are trained on local data, accountable to local regulatory frameworks, and genuinely optimised for local problems, not retrofitted versions of tools built for users in San Francisco or Stockholm. It means pushing for multilateral AI governance at the United Nations that gives developing nations a real voice in setting the rules, rather than inheriting frameworks written by the countries that got there first.
The student in Nairobi deserves an AI tutor as good as the one in Oslo. Getting her that tool requires $50 billion and a great deal more, it requires a fundamental rethinking of who the digital economy is built for, and who gets to shape what it becomes.
*Chloe Maluleke
Associate at BRICS+ Consulting Group
Russian & Middle Eastern Specialist
**The Views expressed do not necessarily reflect the views of Independent Media or IOL.
** MORE ARTICLES ON OUR WEBSITE https://bricscg.com/
** Follow @brics_daily on Twitter for daily BRICS+ updates and instagram @brics_daily