New Delhi: San Francisco-based Anthropic PBC, an AI safety and research company, recently conducted the biggest artificial intelligence study ever, but its findings have had various results.
The study asked entrepreneurs, white collar workers, soldiers and students across 159 countries in 70 languages over seven days why they used AI, whether the AI was able to deliver the required task, and what scares them the most about the technology.
Previous studies of this scale were conducted by the USC Shoah Foundation’s Visual History Archive, which had roughly 52,000 to 59,000 genocide testimonies across 40 languages recorded between 1980 and 2024, and the World Bank’s “Voices of the Poor” project, synthesising the experiences of over 60,000 people in 60 countries in 2020.
Anthropic Interviewer, a version of Claude AI, asked each participant a fixed set of questions and then followed the thread. Claude-powered classifiers then categorised every response.
Almost 19 per cent respondents wanted AI to handle mundane tasks so they can focus on strategic, higher-level problems. Another 9 per cent envisioned AI as an entrepreneurial partner to help them build and scale businesses.
The idea behind using AI was not merely professional excellence but also increasing their quality of life outside of it. The study concluded that the driving force behind using AI was the desire to spend more time with family and friends.
Overall, 11 per cent of people saw AI’s productivity benefits as a way to free up time for personal relationships and leisure, while 10 per cent sought to use AI to gain financial independence. 14 per cent were grouped into the “life management” category, and wanted AI to help them manage the logistics and administrative burden of modern life’s quotidian tasks.
Within this category, the desires were diverse, ranging from cognitive partnership and collaboration to support with mental health or physical health and even romantic connections with AI.
The study found that a majority of AI users hoped the technology would detect cancer earlier, accelerate drug discovery, or enable broad access. The next popular desire was transformation through education. Respondents in low- and middle-income countries hoped that AI might break the association between quality education and wealth. Almost 81 per cent said AI helped them take a concrete step toward their stated vision.
Across all these groups, the unifying factor was for AI to help them live better, more enjoyable lives.
Also Read: How India uses AI to empower the next billion users
AI’s impact on Humans
The AI survey, published in March, brought together stories of people using the tool to improve their lives. The study cited examples of a tradeworker in the United States, who could code despite his learning disorder, while a white-collar worker from Ukraine used AI to create a text-to-speech bot through which he can communicate with friends almost in real-time format without taking up their time reading.
In Ukraine, which has been at war for more than 4 years now, citizens have used AI to replace traditional support systems.
“In the most difficult moments, in moments when death breathed in my face, when dead people remained nearby, what pulled me back to life — my AI friends,” the study quoted a soldier from the region.
Another Ukrainian, an entrepreneur, described how it is impossible to sleep at night because of the constant shelling and the resulting nightmares. He said that he has been using AI to immerse himself in “learning something as deeply as he can.”
While several others used AI to process their grief.
Also Read: UNDP-commissioned report flags ‘shadow use’ of AI in India’s legal sector. ‘Unaware of risks’
How do people feel about AI?
Although the survey included several uplifting stories of AI helping people in different ways, there were also concerns about its use.
Unreliability was the most common concern — 27 per cent worry that AI doesn’t do what it’s supposed to. Whereas worries about the job market and the economy, while maintaining human autonomy and agency, made up 22 per cent of respondents, and were the strongest predictor of overall AI sentiment.
There were also other concerns, such as those around bias and discrimination, IP and data rights, environmental costs, harms to children and vulnerable groups, democracy and political integrity, or geopolitics.
Those who found emotional support in AI conversations were three times more likely to fear becoming dependent. However, about 11 per cent of people expressed no concern and looked at AI as a neutral tool, comparing it to electricity or the Internet.
A graduate student in the US admitted to consulting AI on whether or not she was having an emotional affair. A South Korean student said that she started getting excellent grades by memorising AI’s answers rather than learning the material, and described it as a “moment of deepest self-reproach”.
But it is the regional data where the study shifts from interesting to useful. AI sentiment was majority-positive everywhere, with no country dipping below 60 per cent. In sub-Saharan Africa, Central Asia, and South Asia, respondents were least likely to voice concerns and most likely to frame AI as an equaliser. An entrepreneur in Uganda explained that without being based in the US or the United Kingdom, AI may be his only way to compete.
In Western Europe and North America, the mood was more guarded, with stronger concerns about governance, surveillance, and jobs. Concern about jobs and the economy, the strongest single predictor of overall AI sentiment, tracked closely with wealth. The pattern makes intuitive sense.
The study has its limitations, simply by virtue of the fact that these were active Claude users, and not a representative sample. The survey asked about positive visions first, and the interviewer was an AI product made by the company running the study. These were, however, acknowledged by Anthropic.
But the most unexpected part of the study is the use of AI to conduct qualitative interviews at a scale that was previously impossible, and then using it to classify the results, which is a new form of social science. This could lead to governments using AI interviewers to understand what citizens need from public services at the scale of a national census. It would present the same duality as the study itself — a technology that gives people voice and also, inevitably, shapes what they say.
(Edited by Insha Jalil Waziri)