Who is most optimistic about AI – and who is least?

Anthropic’s report, which surveyed nearly 81,000 people in 159 countries, found that people in sub-Saharan Africa and Asia are more optimistic about artificial intelligence than people in Western Europe or North America.

The survey, published on Wednesday, found that economic benefits from the use of AI were the main aspiration of most respondents, but analysts also warned that not everyone would benefit equally.

Humanities researchers invited users of the Claude large-scale language model to participate in a conversation centered around questions about their usage habits, hopes and fears for AI development.

These conversations were conducted using an Anthropic Interviewer (a variant of Claude trained to conduct interviews) and then analyzed with Claude as well. It starts by filtering responses that are “spammy, disingenuous, or extremely minimal” and then categorizes and tags responses by emotion.

Respondents reported that they have the highest expectations for and see the greatest benefits of AI in the workplace.

According to the report, 18.8% of respondents wanted “professional excellence” in the use of AI. Similarly, 32% reported that AI would be most helpful in improving productivity.

According to Anthropic, most of the productivity gains are related to respondents being able to outsource more mundane tasks and “focus on strategic, higher-level issues.” Some say AI will free them up to do more than just work.

These sentiments were not surprising, as some analysts said that the current stage of AI development is suitable for simpler applications.

“Right now, AI is best suited for use cases that are repetitive, focused, and goal-oriented… similar to specific tasks on an assembly line,” said Lian Jye Su, principal analyst at Omdia, in an email to CNBC.

More specifically, Seema Shah, vice president of insights at market intelligence firm SensorTower, said in an email to CNBC that these applications often include administrative tasks such as “human resources, billing, and other back-office functions.”

According to Anthropic, the economic benefits of AI also seem to favor the entrepreneurial class, with independent workers, including entrepreneurs, small business owners, and people with second jobs, being more than three times as likely as salaried workers to be economically empowered by the use of AI.

However, recent developments have also shown that ostensibly higher-order work may be vulnerable to many of the same disruptions.

After Anthropic launched Cowork (a variant of Claude that can handle more complex tasks such as financial modeling and data management) in February, stocks of companies ranging from software to research firms fell sharply as investors spooked at the impact of these launches.

As companies like Anthropic and Alibaba invest billions of dollars in agent AI, developing models that can autonomously perform actions with limited user supervision, predicting how professional lives will be disrupted may become even more difficult.

“These agents are going to perform increasingly sophisticated tasks on behalf of people, and that’s going to have a huge impact,” Mark Einstein, research director at Counterpoint Research, said on a call with CNBC.

Given the uncertainty with which future developments are expected to further change human work, concerns about turnover emerged as one of the main concerns in Anthropic’s survey, with 22.3% of respondents citing job concerns as their top source of anxiety.

Anthropic’s December 2025 report found that these eviction concerns were “fairly evenly spread across occupations.”

“When I code now, I’m no longer a creator, I’m almost just an observer. I find that I may not even be needed in the role of observer,” Anthropic quoted an anonymous software engineer in the US as saying.

As the development of AI continues at a dizzying pace, analysts are divided on who will truly benefit from the promise of economic empowerment that AI brings.

“I see AI as the great equalizer,” Einstein said. “One of the great things about AI is that in rural Indonesia or Brazil,[people]can access the same AI that they have in the United States or Japan.”

Claude users in emerging economies such as sub-Saharan Africa and Latin America appear to be 10 to 12 percentage points less likely to express negative feelings toward AI than users in Western Europe or North America.

Respondents in sub-Saharan Africa also expressed a stronger desire for entrepreneurship and economic independence through the use of AI than their North American counterparts. Similar differences emerged when North American users were compared to Latin American and Asian respondents.

But while these findings may reflect real-world perceptions of opportunities, particularly related to the use of AI as a mechanism for access and economic mobility, this view of the data is also undermined by research methodology, said Lia Raquel Neves, founder of ethics consultancy EITIC.

While the 80,508 responses that met the researchers’ quality standards was a large sample by any measure, Anthropic was upfront about the methodological limitations associated with conducting independent research on AI from a pool of existing users.

Antropic writes in an appendix that the group of respondents is “skewed toward those who see enough value in AI to continue using it, and are likely to have a more positive vision than a general population sample would produce.”

Almost half of all respondents were also from North America and Western Europe.

“This result should be interpreted as an indicator of how early active users are constructing their AI experiences in different contexts, rather than as a unified picture,” Raquel Neves said in an email to CNBC.

Users in emerging countries appear to be most excited about the prospect of economic benefits from the use of AI, but it remains unclear how evenly the spoils of AI development will be shared.

In a 2025 report, the United Nations Development Program warned that future AI developments could exacerbate existing socio-economic inequalities. Economic benefits tend to be disproportionately accrued by societies with greater capacity and access to digital infrastructure, which often means wealthier countries.

“If the right conditions aren’t in place,[AI]can amplify existing vulnerabilities through digital exclusion, algorithmic bias, and reliance on external systems,” Raquel Neves told CNBC.

While it may be too early to tell who is most likely to lose in the AI ​​race, there is little doubt who will emerge as the winner.

“Whoever succeeds in introducing the (AI) agents that we all begin to use will definitely win,” Einstein said.

Antropic did not respond to CNBC’s request for comment.

—CNBC’s Dylan Butts contributed to this report.