The 2026 estimate and what it represents

The International Telecommunication Union (ITU) estimates that 6.3 billion people will use the Internet in 2026, or 76% of the world’s population. That puts the share above three-quarters of humanity, while leaving about 2 billion people offline, according to the same estimate.

The figure comes from Measuring digital development: Facts and Figures 2026, the ITU’s annual series of connectivity estimates by region and income group. It is an estimate for the year, not a real-time headcount or a measure of every active connection on a given day.

The figure helps measure global Internet use and track its overall trajectory. The ITU reports network coverage and service affordability as separate indicators.

What the income gap reveals

The ITU estimates that 93% of people in high-income countries use the Internet, compared with 32% in low-income countries. A high global penetration rate, in other words, does not mean connectivity is universal.

The comparison by income level shows a marked divide: the ITU estimates Internet use at 93% of the population in high-income countries, versus 32% in low-income countries.

The edition also notes that mobile phone ownership exceeds Internet use in most countries. Owning a device does not mean being connected: the gap between these indicators is a reminder that effective access depends on more than having a phone.

Estimated Internet use in 2026

ITU estimates show the global share online and the contrast between high- and low-income countries.

Estimated Internet use in 2026 · Share of population, except global total
Worldwide6.3 billion people · 76%
High-income countries93%
Low-income countries32%

2026 estimate · Global and country income groups

Estimates describe Internet use, not network coverage, affordability, or connection quality.

Sources and references1

The difference between use and effective access

The 2026 edition presents an estimate of Internet use, alongside separate sections on network coverage and service affordability. It also includes indicators on the gender gap, use among young people, urban-rural differences and digital skills.

The ITU treats Internet use, network coverage and service affordability as distinct indicators in its 2026 edition.

The report’s structure avoids reducing digital development to a single figure: it separates Internet use from network coverage and service affordability. When assessing a public policy or connectivity plan, the practical question is which barrier each indicator addresses: limited coverage, high costs, device availability or the skills needed to use services.

What the 2025 figures show

The ITU’s 2025 edition put Internet use at 74% of the world’s population, or about 6 billion people, up from 71% in 2024. The 2026 edition estimates that 6.3 billion people, or 76% of the global population, use the Internet.

In 2025, the ITU estimated Internet use at 94% of the population in high-income countries and 23% in low-income countries. For 2026, its estimates are 93% and 32%, respectively.

The global figure alone does not establish that any particular policy or technology caused a change in Internet use.

Geographic disparities still matter

The ITU’s 2025 data also show differences between urban and rural areas: 85% of the urban population was online, compared with 58% of the rural population. The ratio between the two rates had narrowed from 1.6 to 1.5 over four years, but convergence was modest.

The geographic gap was especially wide in Africa, where the urban-to-rural usage ratio was 2.6, compared with 1.1 in Europe. In low-income countries, just 14% of the rural population used the Internet in 2025. The global average does not show how access is distributed within each country or region.

To interpret the 2026 estimate, it helps to distinguish three questions: how many people use the Internet, who remains offline, and how good or affordable the service is. The first measures scale; the other two help identify where action is needed. A rising global percentage is progress, but it does not replace analysis of persistent gaps.