People in English-speaking countries including the UK, US, Australia and Canada are more nervous about the rise of artificial intelligence than those in the largest EU economies, where excitement over its spread is higher, new research suggests.
A global split over what has been dubbed “the wonder and worry” of AI appears to correlate with widely divergent levels of trust in governments to regulate the fast-developing technology.
The polling of 23,000 adults in 30 countries, shared exclusively with the Guardian by Ipsos Mori, also showed a quarter of people globally still do not have a good understanding of what AI is, despite it being widely described as the most transformative technology in decades.
On Wednesday, Abba’s Björn Ulvaeus revealed he waswriting a musical with the assistance of AI, describing it as “like having another songwriter in the room with a huge reference frame”.
Britons appear to be among the world’s most worried people about the rise of AI, with two-thirds of people in Great Britain saying they are nervous about the technology being deployed in products and services, and less than half trusting the UK government to regulate AI responsibly.
By contrast half or less than half of people in France, Germany and Italy said products and services using AI made them nervous.
“In the Anglosphere (US, Great Britain, Canada and Ireland and Australia) there is much more nervousness than excitement,” said Matt Carmichael, a senior vice-president at Ipsos Mori. “In European markets we see less nervousness, but also just a mid-range of excitement. Some markets are much more positive than nervous, especially in south-east Asia.”
Only Americans, Japanese people and Hungarians trust their governments less to regulate AI than Britons. The UK government recently delayed a bill intended to regulate AI companies in order to align itself with the stance of Donald Trump’s administration in the US.
Trust in government regulation is lowest in the US, where the president’s election campaign was bankrolled by Silicon Valley technology oligarchs including Mark Zuckerberg, Elon Musk and Jeff Bezos and he recently proposed a bill preventing new state-led regulations of AI.
By contrast last June, the European Union passed the bloc-wide EU AI Act, which bans AI that poses an “unacceptable risk”, for example, systems used for social scoring, and requires systems to declare when AI has been used to manipulate or generate content.
People in India, where the use of misleading AI-generated deepfake videos marked last year’s general election campaign, are also among the most nervous about AI being used in products and services.
The polling also revealed widespread opposition to AI’s use in creating news articles, films and adverts but an equal acceptance that AI will become the primary producer of these things anyway.
The highest levels of excitement about AI were found in Indonesia, Malaysia and Thailand where levels of trust in government regulation were also highest. Polling in those countries was only representative of the more “connected” urban and educated populations but it showed almost double the levels of excitement as in the whole populations of the US and Great Britain.
People in Great Britain were among the most pessimistic about how AI will worsen the job market, with nearly a third fearing AI will replace them entirely at work.
Globally, just 31% of people think the job market in their country will improve because of AI and 35% think it will get worse. But perception of its impact varied widely. Nearly three-quarters of people in Thailand believe it is very or somewhat likely that AI will replace their current job in the next five years, compared with only 14% who believe their job will go in Sweden and one in four in the US, Great Britain and Australia.
Across all 30 countries, the polling showed very few people want AI created-online news articles, films or adverts, but most people think it is likely AI will become the primary producer of all of these things as well as making television programmes, screening job adverts and even creating realistic sports content such as tennis matches between AI-generated players.
Carmichael said this could play out either with increasing public acceptance as AI-generated content becomes more widespread or alternatively a “backlash”.
Some of that resistance is currently being seen with the campaign by musicians in the UK, including Kate Bush and Elton John, for greater protections against copyright infringement by technology companies building large language models (LLMs). There have also been lawsuits in the US where novelists from John Grisham to Ta-Nehisi Coateshave been suingOpenAI and Microsoft for copyright infringement.