People and countries with lower incomes most wary of potential harms from AI and personal data misuse
The 2021 World Risk Poll found that people living in low-income and low-middle-income countries are more likely to say AI would ‘mostly harm’ people in their country in the next 20 years. Likewise, views on personal data use also varies by income status. Both findings suggest that people’s level of financial security plays a role in attitudes towards AI and personal data misuse.
Worry about AI across low-income regions
More than a third of people worldwide who are living in low-income (35%) and lower-middle-income countries (34%), say AI would ‘mostly harm’ people in their country. Specifically, four regions where the majority of countries are classed as low or lower-middle income – Central/Western Africa, Northern Africa, Eastern Africa, and Southern Asia — saw the highest levels of worry about harm from AI, with people in all four more likely to say it will harm than help. In Eastern Africa, more than half (51%) of people said AI will mostly harm people in their country.
These findings suggest that in some countries, the use of AI may be seen as a means of reinforcing and amplifying existing global biases and discriminatory power structures, causing them more harm than good.
It’s a vast contrast to Eastern Asia where optimism about AI was the strongest – 57% of people said AI would mostly help people, while just 13% said it would mostly harm them. This may be unsurprising considering several countries in the region — including China, Japan, and South Korea — are considered leaders in developing AI technologies.
As we have seen in other areas of the World Risk Poll, the perception of risk can also cause harm in itself by excluding already marginalised groups – such as those on low incomes – from the social and economic opportunities that digital technologies offer.
To address the risk of exclusion, it is critical that policymakers, businesses and digital innovators construct and regulate personal data use and AI technologies in ways that counter biases, rather than causing harm by amplifying them, and that they engage with vulnerable communities to give them the confidence to engage with data programmes and technologies that can benefit them.
Dr Sarah Cumbers
Director of Evidence and Insight, Lloyd’s Register Foundation
Concern about personal data misuse
Amongst internet users worldwide who say they are ‘living comfortably’ on their current income, 35% are very worried their personal information will be stolen, compared with 56% of those who are ‘finding it very difficult’ on their current income.
Four in five people in Malaysia (82%) who are finding it difficult to get by financially are worried about their data being stolen. This is the highest rate globally, followed by Indonesia (80%), Poland (69% – the highest in Europe), and then Brazil (65%).
At the other end of the scale, only 5% of people in Lithuania who say they are finding it very difficult to get by on their current income are worried about their personal data being stolen online.
People in Eastern Africa worry most about harm from AI.
More than half of people in Eastern Africa said AI will mostly in their country.
Less optimism about AI amongst religious groups
In addition to financial security, the Poll also found that people who say religion is important in their daily lives express less optimism about AI. Overall, 32% of those who say religion is important in their daily lives feel AI would mostly help people in their country, versus 48% of those who do not.
Though the United States is a global leader in AI, just 32% of people in the United States believe AI will mostly help people in the next 20 years, while 34% say it will mostly cause harm.
The importance of religion to many Americans may factor into this difference, as it is associated with scepticism about the impact of AI. More than half (58%) of Americans say religion is an important part of their daily lives, versus 31% in Northern/Western Europe and only 17% in Eastern Asia. Amongst more religious Americans, only 25% say AI will mostly help people, compared with 43% of Americans who say religion is not an important part of their daily lives.
People in Eastern Asia have confidence in AI.
Optimism about AI is strongest in Eastern Asia where 57% of people said AI would mostly help people.
Those who said AI will ‘mostly help’ versus ‘mostly harm’ people.
Do you think artificial intelligence will mostly HELP or mostly HARM people in this country in the next 20 years?
You might also be interested in...
With the web becoming an increasingly monitored space, each of us has a role to play in safeguarding online privacy
Eastern Africa is not ready to accept artificial intelligence
South Koreans most positive about impact of AI according to new research
Scandinavians optimistic about artificial intelligence – unless they’ve experienced discrimination
Are we right not to trust autonomous vehicles?
