Those are some of the conclusions from a wide-ranging, 25-country survey conducted by the Pew Research Center exploring public sentiments about artificial intelligence.
The headline figures can be read as a tale of wariness, uneven belief in regulators and wide gaps in awareness that echo economic, political and cultural disparities.
Not Excitement, but Concern Is the Global Norm
Half of adults in the United States and Italy say they feel more concerned than excited about AI’s increasing presence in daily life. Another 38% are both concerned and excited in about equal shares, while just one-in-ten (10%) say they are mostly excited. And in Canada, 9% are mostly excited — the lowest percentage among the nations surveyed.
At the other end of the spectrum, Israel is the country that feels most excited among those surveyed, with about 29% feeling mostly enthusiastic. South Korea and Sweden (both 22%), and Nigeria (20%) trail slightly behind. Even then, those “mostly excited” shares are minority positions — underscoring that around the world, caution is the default mode.
What’s driving the unease? Although Pew did not report reasons for their concerns in this study, those most often named in other research and policy venues are:
- Misinformation and deepfakes
- Embedded bias
- Job displacement
- Data privacy (possibly the number one concern overall)
- Cybersecurity
- The environmental cost of large-scale computing
The Paradox of AI Use and Anxiety Among Users
Despite apprehension, usage keeps climbing. About 800 million people use ChatGPT each week, OpenAI says, which would make it among the world’s most-visited websites. That split — high engagement from a position of high concern — seems to indicate that people care about tangible productivity and knowledge access even as they worry about the broader social effects of technology.
There is also a “chronically online” effect, Pew notes — people who are almost always on the internet have significantly more enthusiasm for AI.
Experience close at hand may moderate fear or, failing that, focus the mind on where the benefits lie in fact as opposed to speculation.
Public Trust Shows AI Regulations Vary Widely
Public trust in who writes the rules for AI is divided. Indians are most optimistic in their own country governing AI well with 89% of people expressing confidence, followed by Indonesians at 74% and Israelis at 72%. Greeks are the least trusting (22%).
Americans are largely divided on the matter: 44% say they trust their country to regulate AI responsibly and 47% do not.
Pew detects a partisan split as well, with GOP and right-leaning voters more likely to say they have confidence (54%) than their Democratic and Democratic-leaning counterparts (36%).
Other respondents also had faith in the European Union’s approach, which may be helpful in consideration of high-profile development pieces such as the EU AI Act that imposes risk-based obligations on developers and deployers. The OECD’s and UNESCO’s global frameworks, and multilateral initiatives such as the G7’s AI process, have also influenced what those guardrails might look like if they were put into practice.
Which Countries Feel Most Familiar and Informed About AI
Acquaintance correlates with national income too. About half of adults in Japan, Germany, France and the U.S. say they’ve “heard a lot” about AI. By contrast, that number is 14% in India and 12% in Kenya. Awareness gaps matter: People who know more about A.I. tend to have more complex views of the risks and benefits involved.
There’s a gender divide too. Men are more likely to have heard a lot about AI than women in all but 11 of the countries that Pew polled. Women, meanwhile, are far more likely to say they are most worried about its spread. But usage patterns are complicated; OpenAI has said that more than half of ChatGPT’s users are women — suggesting a familiarity gap that may close as tools like it become embedded at school and work.
Generation also matters. Young adults are more AI-aware pretty much everywhere, but knowledge does not mean hope. And another study by EduBirdie similarly found that many Gen Zers feel like they rely too much on AI — and a surprising portion see the current moment as particularly hard — illustrating that the generation of digital natives isn’t universally bullish about automation.
Why Countries Diverge in Their Attitudes Toward AI
National context shapes sentiment. There is somewhat higher enthusiasm in places with robust tech sectors and visibly increased productivity (like South Korea and Israel). Countries that have also fought polarizing online misinformation or labor market angst tend to report more concern too, including in the United States and parts of Europe.
Policy clarity helps. To the extent that governments articulate rules and plans for enforcement based around risks, people might feel more confident that AI’s upside can be seized without losing sight of safety or rights. On the flip side, regulatory uncertainty or institutional gridlock tend to be proxies for skepticism.
What to Watch Next as AI Adoption and Policy Evolve
Three signposts will decide whether concern recedes or grows:
- Visible progress on safety standards and enforcement
- Clear evidence that AI creates opportunity without depressing wages or compromising privacy
- Plausible guardrails against deepfakes and election interference
If policy makers and industry do deliver on those fronts, public opinion could move from uneasy to cautiously optimistic. If not, the current global stance — curious but worried — could harden into opposition.
