Artificial Intelligence: Hype for the few, harm for the many?
Who is AI really for?
The artificial intelligence (AI) boom is sweeping the globe, with big tech and their investors pouring staggering sums into AI infrastructure. In 2025, global AI spending is projected to surpass $500 billion, driven by tech giants like Amazon ($110 billion), Microsoft ($80 billion), and initiatives like Project Stargate ($100 billion). But in a world where 4.8 billion people—over half the global population—are poorer than a decade ago, this AI gold rush raises a stark question: who is AI really for?
Drawing on insights from Nobel Prize-winning economist Simon Johnson (listen to our Equals podcast), we unpack how AI could widen divides unless governments act to steer it toward the public good.
AI, in numbers
Getting priorities straight. As inequality skyrockets, with billionaires $3.3 trillion richer since the pandemic, the promise of AI as a tool for humanity feels hollow. A mere $35 billion—less than 7% of 2025’s AI spending—could end world hunger and ease the debt crisis strangling Global South communities. Yet, the AI race is led by a handful of powerful players, mostly men in Silicon Valley, shaping a technology that risks deepening inequality and accelerating the climate crisis.
AI’s unequal blueprint. AI’s development is concentrated in the hands of a few. Silicon Valley attracts 65% of global AI-native funding, double its share of overall tech investment. Over 80% of AI and IT leaders are white men, embedding their priorities into the technology’s design. The top 100 companies, mostly in the US and China, account for 40% of global corporate AI research and development (R&D), with market values rivalling entire continents’ GDPs (Africa’s). This elite control means AI is built for profit, not people, serving markets that cater to the richest rather than addressing pressing societal needs like poverty or hunger.
Bias baked in. AI isn’t neutral—it mirrors the biases of its creators. The Gender Shades study revealed facial recognition systems had error rates as low as 1% for light-skinned men but up to 34.7% for darker-skinned women, due to unrepresentative training data. These biases, rooted in datasets and algorithms shaped by a narrow demographic, perpetuate discrimination based on gender, race, and ethnicity. Without diverse voices in AI’s design, it risks entrenching systemic inequities, from hiring algorithms that favour men to surveillance tools that disproportionately target marginalised groups.
Job displacement. AI’s threat to labour is real. The International Monetary Fund (IMF) reported that AI is likely to affect 40% of jobs globally. In advanced economies, this figure rises to 60%, with about half of those jobs potentially experiencing negative impacts through lower salaries, reduced hiring, or outright disappearance. Workers in repetitive tasks are particularly vulnerable to displacement.
The Global South left behind. The gap between the Global North and Global South risks widening further. AI requires robust digital infrastructure, including high-speed internet, reliable electricity, and data storage. Many regions in the Global South lack these foundational enablers. For example, Sub-Saharan Africa had an internet penetration rate of only about 37% in 2023. Compounding this, over 600 million people in Sub-Saharan Africa still lack access to electricity. Without such basic access, developing countries cannot effectively adopt or leverage AI solutions. If AI's promised benefits in health and education are to materialise, access and control must become far more equitable. Otherwise, AI will exacerbate global divides, with richer countries monopolising benefits like precision medicine or climate modelling.
Fuelling the climate crisis: AI’s environmental toll is staggering. Training large models and running data centres guzzle energy, often powered by fossil fuels. A single hyperscale data centre can consume 876,000 MWh annually, equivalent to the electricity use of 4.8 million people in Sub-Saharan Africa. Cooling these facilities demands vast water resources. For instance, a single hyperscale data centre can average around 2 million litres of water per day. Google's data centres, which power a significant portion of global AI, used approximately 21.2 billion litres in 2022. To put this in perspective, this amount is roughly equivalent to the annual water needs of about 2 million people in Sub-Saharan Africa at a basic daily consumption rate, highlighting the significant environmental footprint of these operations. Even with renewable energy, the breakneck pace of AI expansion outstrips sustainable limits, contributing to greenhouse gas emissions and straining water-scarce communities.
Political and economic capture: AI isn’t just a technology—it’s a political and economic force. Tech moguls like Elon Musk, Jeff Bezos, and Sam Altman wield growing influence, lobbying governments to prioritise profit over regulation. The “AI race” is framed as a matter of national security and prosperity, dampening political will to enforce accountability. This concentration of power risks “political capture,” where a handful of billionaires shape AI’s future to serve their interests, not society’s. The result? A world where the richest 1%, who already own 43% of global financial assets, tighten their grip. The problem is not the robots, but who owns the robots.
A force for good? In the right hands, with strong regulation and broad-based ownership, AI offers real hope. It could support new scientific and medical breakthroughs, improve our understanding of climate change and the natural world, improve health and safety for workers and drive efficiency.
The bottom line: The hype around AI is deafening, but its current trajectory so far serves the super-rich, not the 99%. We’re racing toward a future where a handful of tech titans dictate the rules, leaving the planet and its people to pay the price. Governments, activists, and communities must act now to wrest control of AI from the billionaires and design it for the common good.
Something to read/listen to
Listen to Nobel Prize winner and MIT professor Simon Johnson dive deep into the world of Artificial Intelligence (AI) on the Equals podcast.
Read these articles in the International Journal of Communication and the London School of Economics about how inequality research misses the talking about “inequality as spectacle”. It’s not just about stats and theory but stories that shift mindsets.
Read about the International Trade Union Confederation’s 2025 Global Rights Index, on how workers’ rights across the world are in decline. Covered well by the Guardian.
Read about the launch of the Global Alliance Against Inequality at the Hamburg Sustainability Conference.