Welcome to the new digital divide

In 2019, GZERO reported on a brewing backlash against AI’s facial recognition and surveillance capabilities. Seven years later, opposition to the technology, which has advanced dramatically since then, is intensifying. Asurvey of more than 18,000 people in 15 countries found that over one-third ranked AI development among their top concerns over the next five years, ahead of immigration, energy security, and geopolitical rivalry.

But attitudes toward AI vary sharply around the world. Ipsos polling found that people in Asia and Latin America are more likely to believe AI’s benefits outweigh its risks and to express enthusiasm for AI-powered products and services. By contrast, North Americans are far more likely to say AI makes them nervous and are less convinced that its benefits justify its costs.

North America says “not in my backyard.” The AI backlash in North America centers heavily on environmental costs. The data centers needed to train and operate AI models consume enormous amounts of electricity and large volumes of water for cooling. A large data center uses up to 5 million gallons a day, equal to the daily needs of a town of 10,000 to 50,000 people.

In Canada, a recent survey found that68% of respondents would oppose a large AI data center near their home, citing concerns about higher energy costs, environmental damage, and neighborhood quality.Seven in ten Americans hold similar views. And politicians are responding: Mississauga, Canada’s seventh-largest city, voted last month totemporarily pause new AI data center construction. In the US, New Yorkimposed a one-year moratorium in July on permits for large data centers, and voters of Monterey Park, California, approvedan outright ban in June.

In the US, data centers have become a political issue ahead of the November midterms. Democratic Senate candidate Abdul El-Sayedcriticized plans for a data center in Michigan, while fellow Democratic House candidate William Lawrence made opposition to four local data centers a key plank in his platform after public outcry.In Ohio, both Republican nominee for governor, Vivek Ramaswamy and Democratic candidate for governor Amy Acton have both pledged to restrict data centers.

Fears of job losses also fuel opposition in North America. A recent Gallup survey found that among Americans ages 14 to 29, excitement about AI fell 14 percentage points over the past year, to 22%, while anger rose nine points, to 31%. Two factors appear to be driving that concern. First, North America has large white-collar sectors, including law, finance, and media, where generative AI is automating parts of entry-level work. Second, the US is home to many companies building AI systems, which gives citizens greater exposure to news about AI and its disruptive potential.

But so far, those losses appear limited: Research from Stanford University found that employment in occupations most exposed to AI fell by 0.2% year over year in June 2026, while employment in the least-exposed occupations grew by 0.6% over the same period.

Asia and the Global South embrace AI. By contrast, polling suggests people in parts of Asia and Latin America are far more enthusiastic about AI. Overall, 64% of respondents across Asia Pacific and 56% of respondents in Latin America say that using AI products and services excites them, compared with just 29% in North America.

Why the difference? Opportunity. A recent study by the World Bank found that generative AI threatens far fewer jobs through automation in low- and middle-income countries (about 4.5%) than in advanced economies (over 14%), because fewer people in those countries work in knowledge-intensive professional services.

Despite the controversy surrounding them in the West, many developing countries—including somevery small states, as GZERO reported last week—welcome data centers as a means of economic development. Although countries such as Thailand and Indonesia are concerned about environmental impacts, 69% and 65% of their citizens, respectively, believe that AI’s benefits outweigh its environmental costs. In India, 71% share that view, although activists recently protested Google’s planned $15 billion AI data center complex in Visakhapatnam, the largest city in the Indian state of Andhra Pradesh. Marching under the banner “We cannot drink DATA,” they warned that the project could worsen an already fragile local water supply.

China’s all in on AI. China is the world’s most bullish country on AI, with 80% of its population sayingthey are excited about its prospects. Several factors help explain that optimism. Over the last three decades, rapid technological progress has raised living standards and improved infrastructure. China has also built leading chip and electronics industries, which many view as a source of national pride.

Meanwhile, the Chinese one-party system makes it very difficult toorganize political opposition to AI data centers, unlike the democratic systems in the United States and Canada. China also has the highest percentage of citizens—77%—who believe AI’s potential benefits or society outweigh its environmental costs. The country has about one-tenth the number of data centers as the US: 449, compared to 5,427 in the US. As a result, fewer Chinese citizens have been exposed to the negative impacts of those facilities.

The new digital divide. Ultimately, the AI backlash is not about the technology itself, but about its impact. In North America, many people fear that AI will threaten white-collar jobs and damage the environment. In China and parts of the Global South, many see AI as a path to higher productivity and faster growth. The key question is whether those countries can overcome the hurdles to implementation—and whether states with the capacity to deploy AI choose to pause.