Key Takeaways
- The two countries building much of the world’s AI judge it in opposite ways: 83 percent of people in China agree that the benefits of AI outweigh its risks, against 48 percent in the USA, where a narrow majority (52%) now disagrees and strong disagreement (21% across North America) outweighs strong agreement (15%).
- Neither country is an outlier; each represents a wider perspective: China sits with the emerging economies that see AI as a route to development (Vietnam 84%, Nigeria 82%, Kenya and Egypt 79%, India 78%), while the USA sits with a closely divided Europe (Germany and the UK 49%, France and Sweden 50%, Italy 51%, the Netherlands 52%).
- Public acceptance of AI cannot be assumed: This is especially the case in some countries at the forefront of AI development such as the USA, and also some parts of Europe. Building trust will require responding to concerns about jobs, privacy, misinformation, energy use, and accountability.
Who Thinks AI Is Worth the Risk? China Yes, America No
The argument about artificial intelligence has shifted. Two years ago, the public debate was about what AI could do; in 2026, it is about whether the pace of its development should be slowed, with open letters on existential risk, fights over data centre energy and water use, lawsuits over training data, and serious political arguments about AI and jobs. In the middle of that debate, GlobeScan asked people in 33 markets whether they agree that the benefits of AI outweigh its risks. The answers show that AI is being deployed globally but judged locally, and nowhere is the contrast sharper than between the two countries that dominate its development.
In China, 83 percent of people say the benefits of AI outweigh the risks. In the USA, 48 percent agree, and 52 percent do not. That 35-point gap separates the two countries that between them build most of the world’s frontier AI. Globally, two-thirds (66%) agree, but only one in five (20%) agree strongly, with 46 percent no more than somewhat convinced. The world’s majority in favour of AI is wide but thin, and the American position shows how quickly it can tip.
Neither country is an outlier, but is instead representative of group of markets holding a very different perspective on AI. China sits with the emerging economies, where AI is framed as a route to growth and opportunity, including Vietnam at 84 percent agreement that AI’s benefits outweigh its risks, Nigeria at 82 percent, Kenya and Egypt at 79 percent, India at 78 percent, Indonesia and Thailand at 77 percent, and regional agreement of 76 percent in Africa and the Middle East and 74 percent in Asia-Pacific. The USA sits with Europe, where opinion is close to evenly split, including the UK and Germany at 49 percent, France and Sweden at 50 percent, Italy at 51 percent, the Netherlands at 52 percent, against a European average of 52 percent and a North American average of 48 percent. This pattern tracks what people expect AI to do for them. It is welcomed where it is anticipated as development and jobs and tends to be more contested where there is an existing argument about employment, energy costs, misinformation, and corporate power.
WHAT DOES THIS MEAN?
The AI race is often discussed in terms of technological capability, investment, talent, and infrastructure. But these findings suggest social licence may prove equally important. Public acceptance cannot be assumed, even in countries at the forefront of AI development such as the US, and also some parts of Europe. For business and policy leaders, the lesson is that technical progress alone is unlikely to secure long-term success. Building trust will require listening closely to how people experience AI in their daily lives and responding to concerns about jobs, privacy, misinformation, energy use, and accountability. The companies and countries that succeed may not simply be those that build the most powerful AI, but those that retain the public confidence needed to deploy it at scale.
How This Insight Was Generated: This analysis draws on insights from a GlobeScan public opinion survey conducted online among 32,106 respondents across 33 markets in July and August 2026.
Survey Question: For each of the following statements, please indicate if you strongly agree, somewhat agree, somewhat disagree, or strongly disagree. [The benefits of artificial intelligence (AI) outweigh the risks.]
