An unprecedented large scale survey on AI trust levels has been analyzed to offer a 2022 snapshot of global AI sentiments.
In a global survey in Sep/Oct 2022 involving 17,193 respondents from 17 countries (1,001 to 1,021 respondents per country) on matters involving trust levels in AI technology, the following trends were analyzed: AI trust, awareness and acceptance levels; perceptions of AI benefits and risks; perceived guardians of AI development and governance; principles of Responsible AI; and AI trends in the workplace.
The 17 countries were chosen based on three criteria: representation across all nine global regions; leadership in AI activity; and readiness and diversity on the Responsible AI Index: Australia, Brazil, Canada, China, Estonia, Finland, France, Germany, India, Israel, Japan, The Netherlands, Singapore, South Africa, South Korea, the US and the UK.
Here are the qualitative summaries of some country/regional-level findings of this survey, which is touted as a world-first deep dive into trust and global attitudes towards AI:
| Australia | Australian respondents had broadly ambivalent attitudes about AI, tending to rank alongside Western countries such as Canada, the US and the UK. Respondents there had the lowest benefit perceptions of all countries surveyed, with only the Japanese respondents expressing less interest in learning more about AI. |
| China | Respondents there consistently appeared at the top of the rankings — alongside India — in positively perceiving AI. While her respondents are among the least likely to believe AI needs to be regulated, they expressed some of the highest confidence in entities to regulate AI, revealing that they were comfortable with all forms of regulation. |
| Finland | While Finns were the least trusting of AI among all the respondents, they had average acceptance, perceived the benefits of AI, were more likely to express positive than negative emotions about it, and were more likely (compared to most Western countries) to believe that AI will create more jobs than it will eliminate. Finns were among the most knowledgeable respondents about AI in the entire survey. Their difference from other European countries may be attributed to their early AI strategy and investment in public education around AI. |
| France | French respondents were broadly like other Western countries in being wary about AI and reporting a relatively low understanding of it. Over two-thirds of French respondents were fearful of AI, the highest percentage of all countries surveyed. |
| Germany | Germans in the survey had largely positive attitudes about AI applications, consistently ranking at or near the top of Western countries on positive metrics. However, this does not appear to translate to the use of AI at work, and while respondents from there suggested they had comparatively high subjective knowledge of AI, their awareness of AI use in common applications was lower than the average of the countries surveyed. |
| India | India often ranked alongside — and sometimes ahead of — China respondents in having the most positive perceptions of AI. They were unique in that less than half believed AI needs to be regulated, with a majority thinking the appropriate safeguards are most likely already in place. |
| Israel | Israelis respondents had mainly positive attitudes about AI, often more favorable than Western countries, but less so than Singapore and the BICS countries (Brazil, India, China, and South Africa). One aspect in which Israel was very different from all other countries is the respondents’ comparative lack of desire for AI to be regulated by an independent regulator. |
| Japan | Japan ranked amongst the lowest of all countries in their respondents’ perceptions and understanding of AI. They lacked trust, acceptance and awareness, and many were not interested in learning more about AI. They did not believe there were sufficient safeguards around AI: however, a majority still perceived the benefits of AI. Negative attitudes may be linked to the Japanese cohort being the oldest in the sample. |
| Singapore | Singapore respondents consistently reported more positive attitudes towards AI than those of other countries, often clustered with, or just behind, the BICS countries on key measures. They were among the most confident in entities to develop and govern AI, and some of the most trusting of AI at work. |
| South Korea | Almost all South Koreans surveyed here perceived AI as beneficial and most accepted and had positive sentiments about it. However, many perceived it as risky and believed regulation is necessary. This balanced perspective may be partially attributed to most South Koreans in the survey reporting that they understood AI, potentially enabling them to appreciate its expected benefits and challenges. |
| The US | The American respondents were wary about trusting and accepting AI and felt ambivalent about its benefits and risks. While these findings are similar to that of other Western countries in the survey, American respondents differed from their Western counterparts in their beliefs about AI regulation and who should regulate it. Fewer believed AI regulation was necessary than in the UK, Canada, Finland and the Netherlands. Americans respondents were also among the least likely to agree that the government alone should regulate AI, but most supported co-regulation. |
| The UK | Respondents there were ambivalent about AI. While most accepted AI and believe it has benefits, few were willing to trust AI, and many were fearful and worried, expressing concern about its risks. This pattern is similar to that found in several other Western countries, including the US, Canada and Australia. Out of all the countries, the UK respondents were most likely to disagree that AI will create more jobs than it will eliminate, while also having among the lowest rates of perceived AI awareness. |
The people at large also need to be confident that these safeguards are enacted, and that AI is designed and used in a human-centric way to help people and support their understanding. According to the most common sentiments expressed. Additionally:
- Four pathways were identified for strengthening the trustworthy and responsible use of AI systems and the trusted adoption of AI in society: institutional; motivational; uncertainty-reduction; and knowledge/public education.
- While proactively investing in these trust foundations can be time and resource intensive, the data suggests it is critical for sustained acceptance and adoption of smart technologies over time, and hence support good returns on investments.
Takeaway lessons
In total, the data analyzed highlighted the importance of developing adequate governance and regulatory mechanisms that safeguard people from the risks associated with AI use, according to the survey report by KPMG.
The people at large also need to be confident that these safeguards are enacted, and that AI is designed and used in a human-centric way to help people and support their understanding. According to the most common sentiments expressed. Additionally:
- Four pathways were identified for strengthening the trustworthy and responsible use of AI systems and the trusted adoption of AI in society: institutional; motivational; uncertainty-reduction; and knowledge/public education.
- While proactively investing in these trust foundations can be time and resource intensive, the data suggests it is critical for sustained acceptance and adoption of smart technologies over time, and hence support good returns on investments.
The report author further noted that, given AI’s rapid and widespread deployment, it is expected to be important to regularly re-examine public trust and expectations of AI systems as they evolve over time — to help ensure AI use is aligned with and meeting changing public expectations.
