ITU warns today's digital divide could become tomorrow's AI divide

Uneven access to AI

Lamanauskas writes that close to 20 per cent of the world's working-age population is already using generative AI. At the same time, about 2.2 billion people remain offline, and advanced AI computing capacity sits in only a handful of economies.

According to the ITU, around 90 per cent of known AI supercomputer processing happens in just two countries. Latin America and Africa together account for only 3.3 per cent of data centre capacity. This is the basis for the warning that "today's digital divide becomes tomorrow's AI divide."

Investment is booming, but not evenly

Citing the latest World Investment Report from UN Trade and Development (UNCTAD), the article says strategic sectors including AI infrastructure and technologies made up 44 per cent of global greenfield investment value in 2025, up from 16 per cent in 2020. Announced investment project values for data centres alone rose by USD 235 billion in 2025.

Low- and lower-middle-income economies, however, attracted only around 10 per cent of strategic-sector greenfield investment between 2020 and 2025.

Three priorities: infrastructure, people and trust

Lamanauskas says the answer must go beyond building ever more powerful AI systems, and sets out three priorities.

• Infrastructure: Connecting everyone meaningfully by 2030 will require around USD 1.6 trillion in physical infrastructure investment, rising to as much as USD 2.8 trillion when affordability, skills and policies are included. He also notes that data-centre electricity demand is expected to roughly double by 2030.
• People: The World Economic Forum expects nearly 40 per cent of workers' core skills to change by 2030, yet according to UNCTAD, fewer than one-third of developing economies had AI education or training strategies in place in 2025.
• Trust and standards: The number of countries with a national AI strategy or policy has risen from four ten years ago to 104 this year, but AI is advancing faster than many countries can build governance frameworks. International standards, he writes, help technologies work together, reduce costs and make AI systems safer and more reliable, particularly for smaller businesses.

What ITU says it is doing

The ITU describes itself as the United Nations agency for digital technologies, with 194 Member States and more than 1,000 members from industry, academia and civil society. Lamanauskas lists a series of initiatives:

• The AI for Good Global Summit, which brought together more than 12,000 participants from 177 countries this year.
• The UN Global Dialogue on AI Governance, supported jointly by ITU and UNESCO.
• The Digital Infrastructure Investment Catalyser, launched with UNCTAD and eight multilateral development banks, bringing together over 100 organizations to turn digital infrastructure needs into investable projects.
• A forthcoming Digital and AI Infrastructure Investment Playbook, meant to help assess infrastructure and computing needs and identify financing gaps.
• Work at the World Bank's Global Digital Summit on project opportunities in Bangladesh and Pakistan.
• The ITU Academy, which reached 83,000 learners last year, and the AI Skills Coalition with over 80 partners.

On standards, ITU says it has nearly 500 AI-related standards published or under development. The International AI Standards Exchange database, an initiative of the World Standards Cooperation with ISO and IEC, features nearly 1000 AI standards.

What it means

The article frames AI as a question of who benefits, not only how powerful models become. Lamanauskas poses three questions: how developing economies can capture more value from AI, how investment needs can become viable projects and partnerships, and how AI investment can create jobs and strengthen businesses.

He says governments, development banks, industry and international organizations all have a role, and that no single country or organization can deliver the transformation alone. His proposed yardstick is whether a small business can compete, a farmer can prosper, a hospital can deliver better care, and every country has the chance to innovate.