Most people support the need for trustworthy artificial intelligence (AI), which they expect to be regulated, with those in India, China, and Singapore leading the Asia-Pacific in the level of trust they have that current regulations are adequate in terms of keeping the use of AI safe.
Some 97% of respondents worldwide supported the principles for trustworthy AI, with three in four more willing to trust AI systems when assurance mechanisms were in place, according to a KPMG study released Tuesday. Another 71% expected AI to be regulated and 61% said they were wary about trusting AI systems, the study found.
Produced in partnership with the University of Queensland, the report surveyed more than 17,000 individuals across 17 global markets, with a sample size of about 1,000 in each country. Data from the online survey was collected between September and October 2022.
And while 85% of people believed AI would provide a range of benefits, just half believed these outweighed the risks. Some 67% expressed low to moderate acceptance of AI.
In Singapore, though, eight in 10 were accepting of AI, above the global average of 71%. Another 72% in the Asian country expressed optimism about the technology, the study revealed.
Also: How does ChatGPT work?
About 90% of Singapore respondents believed AI offered a range of benefits, with 59% saying these outweighed the risks, the study found. Another 60% were willing to trust the use of AI at work, where 44% acknowledged AI would replace jobs in their work area. Just 27% disagreed AI would create more jobs than it would remove.
Some 53% of people in the country had trust that current regulations and safeguards were sufficient to make its use safe, placing Singapore third across the global study, behind India and China.
The KPMG report noted “significant differences” across nations with regard to public trust in their government involving the use and governance of AI, where half of respondents in the U.S., the U.K., and Japan expressed a lack of confidence.
Some 49% in the U.S. had no or low confidence in the government, while 39% had no or low confidence in commercial organizations to govern and develop AI. In the U.K., 45% and 31% had no or low confidence in the government and commercial organizations, respectively, to govern and develop AI.
In Japan, 47% and 43% respectively had no or low confidence in the government and commercial organizations to govern and develop AI.
Respondents in China and India were most likely to believe appropriate safeguards were in place, at 74% and 80%, respectively, followed by about half in Brazil and Singapore. Their peers in Japan and South Korea were the least convinced, at 13% and 17%, respectively, reflecting similar views of the majority in western countries, the study noted.
Only 30% in the U.S. as well as the U.K. believed current regulations and safeguards were sufficient to ensure AI use was safe, as did 32% in the Netherlands and 39% in Germany.
Across the board, 61% of respondents believed the long-term impact of AI on society still was unclear. This result might help explain why 71% expected AI to be regulated and the majority, with the exception of India, viewed AI regulation as necessary.
However, just two in five respondents believed current regulations and safeguards facilitated the safe adoption of AI, indicating public dissatisfaction with AI regulation, the KPMG report noted.
However, three out of four people expressed more willingness to trust AI when mechanisms were in place to assure ethical and responsible use, such as monitoring systems for accuracy and reliability, independent AI ethics reviews, and codes of conduct.
Awareness of AI appeared high among Asian countries, with China, India, South Korea, and Singapore leading the global pack in terms of understanding of the technology.
China ranked highest in terms of understanding and interest in AI, with 82% saying they understood the technology, and 96% wanting to learn more about it. Some 75% of people in the country used common applications that contained AI and 58% knew when AI was used in such applications.
In the U.S., 55% said they had a low understanding of AI and when the technology was used, although 73% expressed interest in learning more about it. Some 63% used common applications containing AI, but 49% were unaware that AI was used in such applications.
Globally, 82% were aware of AI, but half acknowledged they did not understand the technology and how it was used. Two out of five people were unaware that AI powered common applications that they used.
“[The study findings] highlight the importance of developing adequate governance and regulatory mechanisms that safeguard people from the risks associated with AI use,” the report noted. “The public also needs to be confident these safeguards are enacted and that AI is designed and used in a human-centric way to help people and support their understanding.”
The report continued: “There’s a range of resources to support organizations in embedding principles and practices of trustworthy AI into their operations and putting in place mechanisms that support stakeholder trust in the use of AI. While proactively investing in these trust foundations can be time and resource intensive, this research suggests it’s critical for sustained acceptance and adoption of smart technologies over time and, hence, a returns on investment.”
Singapore most AI-ready
A separate study released Tuesday also placed Singapore in the lead among its Asia-Pacific counterparts in overall AI readiness, scoring 70.1 on the index, which is ahead of 11 other regional economies. Japan and China placed second and third, with a score of 59.8 and 59.7, respectively, according to Salesforce.com’s 2023 Asia-Pacific AI Readiness Index.
Singapore’s latest score was higher than the 65.6 figure it got in the 2021 index, when it also placed first, Salesforce noted.
Currently in its third iteration, the bi-annual index was commissioned by the tech vendor and compiled by Access Partnership. It used a total of 15 indicators to measure AI readiness of both businesses and governments and its impact on socio-economic opportunities.
Seven indicators are used to assess business readiness — specifically, how the private sector is equipped to adopt AI — while eight indicators are used to evaluate government readiness, and how policymakers and regulators are enabling AI through funds and frameworks.
Singapore led in both business and government AI readiness, with a score of 53.6 and 86.5, respectively. The study pointed to the country’s “conducive policy and business environment” etched around its AI-related initiatives, including the Singapore government’s national AI strategy, which outlines plans to develop and roll out AI applications across several key sectors. Its AI Verify Foundation also looks to drive the responsible use of AI via test frameworks and best practices.
China ranked second in business AI readiness with a score of 43.1, followed by South Korea at 42.6, while Australia placed second in government AI readiness at 77.7, followed by Japan at 77.5.
The index, though, revealed a “growing divide” between mature and emerging digital economies. With the exception of Singapore, other Southeast Asian countries — comprising Indonesia, Malaysia, Thailand, the Philippines, and Vietnam — all scored below the Asia-Pacific average of 51.15.
“However, these countries have all launched national AI policies between 2021 and 2022, suggesting it may be a matter of time before we see the impact of the policies on the ground,” Salesforce noted.
Of the 12 Asia-Pacific economies surveyed, five — Australia, Indonesia, New Zealand, Singapore, and Thailand — saw improved overall AI readiness from their 2021 score, which the report attributed to the respective country’s AI-related initiatives that were rolled out between 2021 and 2023.