In 2024, private AI investment in the United States reached $94.6 billion — more than China and Europe combined. The AI revolution is real, but it is not arriving equally. A small number of countries control the money, the systems, and the narrative. The rest of the world watches — and increasingly worries. This story follows the money, the jobs, the machines, the people, and finally the question everyone is asking: will AI help us, or harm us?

Chart 1 — The Money Race: AI Investment Is Surging — But Only in One Place

Insight_1 : US private AI investment grew more than 10× between 2013 and 2024. China peaked in 2021 and has since stalled. Europe barely registers.

Source: Quid via AI Index Report; U.S. Bureau of Labor Statistics (Our World in Data, 2026). Inflation-adjusted USD.

Private AI investment has exploded since 2020, accelerating sharply after the launch of ChatGPT in late 2022. The United States now commands an overwhelming share of global AI capital, driven by multi-billion dollar bets on OpenAI, Anthropic, and Google DeepMind. China, once a close competitor, plateaued after 2021. Europe remains a distant third. This concentration of capital means the direction of AI development — and the values embedded in the systems being built — is decided by a remarkably small number of actors.

Chart 2 — The Jobs Shift: AI Is Reshaping Labour Markets Unevenly

Insight_2 : Singapore leads the world in AI job demand. Australia sits mid-table. Many countries are barely moving — and the gap is widening every year.

Source: Lightcast via AI Index Report (Our World in Data, 2026).

The share of job postings requiring AI skills has grown in every country shown here — but the gap between leaders and laggards is widening. Countries at the top are not just adopting AI tools; they are structurally rewiring their labour markets around them. For workers in countries lower on this chart, the question is urgent: how do we close this gap before it becomes a permanent structural disadvantage?

Chart 3 — The Power Concentration: Two Countries Are Building the Future

Insight_3 : The United States has built more large-scale AI systems than the rest of the world combined. China is second. Every other nation is far behind.

Source: Epoch AI via Our World in Data (2026). Large-scale = training compute > 10²³ FLOP (≈ hundreds of thousands of dollars to train).

The United States and China account for the overwhelming majority of the world’s most powerful AI systems. This is not merely a technological lead — it is a geopolitical one. Nations without frontier AI capability depend on systems built elsewhere, trained on data they did not provide, and governed by rules they did not set. France is a notable European outlier, but the gap remains vast.

Chart 4 — The Human Divide: Age, Gender, and Attitudes Toward AI

Insight_4 : Women are consistently more likely than men to believe AI will harm society — a pattern that widened between 2019 and 2021. Younger people are more optimistic, but less than you might expect.

Source: Lloyd’s Register Foundation World Risk Poll via Our World in Data (2022). 2019 and 2021 are the two available survey waves; 2021 is the most recent data available from this poll.

The generation gap in AI attitudes is real but smaller than headlines suggest. The most striking finding is the gender divide: women are consistently and significantly more likely than men to believe AI will mostly harm society — and this gap widened between 2019 and 2021, precisely as AI tools became more mainstream. This is not pessimism without cause. Women face disproportionate exposure to AI’s documented harms: biased hiring algorithms, facial recognition systems that fail on darker and female faces, and AI-enabled harassment and deepfakes. Meanwhile, AI’s headline gains — productivity tools, high-salary jobs, startup wealth — are concentrated in sectors and roles where men are over-represented. If the people building AI do not resemble the people most harmed by it, the systems they build will reflect that imbalance.

Chart 5 — The Global Verdict: Does the World Believe AI Will Help or Harm?

Insight_5 : East Asia is optimistic. Latin America and Southern Europe are not. The world’s most sceptical nations are also those with the least power to shape AI’s future.

Source: Lloyd’s Register Foundation World Risk Poll via Our World in Data (2022). Top 12 most optimistic and 12 most pessimistic countries shown. 2021 is the most recent survey wave available from this dataset.

The world is deeply divided on AI’s future. East Asian nations express strong confidence that AI will help. Many Latin American, African, and Southern European nations lean toward harm. This is not a simple rich-versus-poor divide — it reflects AI literacy, institutional trust, and lived experience of technology’s benefits and costs. The countries most sceptical of AI are often those with the least power to govern it.