Here’s what the World had to say about the AI economy
AI is already here - and people are choosing to engage with it
AI is already deeply embedded in people's daily lives. Nearly all respondents (98%) encounter AI systems at least weekly, and over half use AI daily in both their personal lives (56%) and at work (54%), with personal adoption slightly outpacing professional use. Yet there is a clear boundary: 57% have never let AI act unsupervised on their behalf, suggesting people are enthusiastic users but cautious delegators.
How often are people encountering AI?
Frequency of AI interactions in the last 3 months — (n=1,045)
FIGURE 1
When AI does act on people's behalf, it clusters around low-stakes cognitive tasks: automating workflows, researching and acting on options, and creating content. A notable 23% have had AI communicate for them. But the drop-off is steep once real-world consequences enter the picture – only 8% have let AI complete financial transactions, and just 10% have used it to submit official documents. A third of respondents say AI has taken none of these actions on their behalf at all. Read alongside the previous finding that 57% have never let AI act unsupervised, a clear pattern emerges: people are comfortable using AI as a tool they direct, but far more reluctant to let it act as an agent with real-world authority – particularly where money, legal identity, or physical wellbeing are at stake.
What action has AI taken on your behalf?
Percentage who used an AI assistant to carry out each action in the past month — (n=1,051)
FIGURE 2
Automated tasks or workflows
29%
Researched options and took action
27%
Created or published content
25%
Communicated on your behalf
23%
Scheduled or booked services
13%
Assisted with accessibility or health
11%
Submitted official documents
10%
Managed work/farm/business operations
10%
Completed financial transactions
8%
Organizad community or group activities
6%
None of the above
33%
0
5
10
15
20
25
30
35
40
At the same time, AI's impact on employment is no longer speculative. Six in ten respondents know someone who has lost a job to automation, and over a third know several people. For most of the sample, job displacement is something they've seen up close. Yet these same respondents are enthusiastic adopters themselves, creating a tension that runs through the rest of the survey: people are actively choosing to use AI while watching it displace workers around them.
How close to home is AI-driven job loss?
"So far, how has your community been affected by job loss from automation?" — (n=1,042)
FIGURE 3
Indeed, the prevailing personal experience of AI remains positive. 72% of respondents say AI has noticeably or profoundly improved their daily lives, and when asked what has changed most, 78% point to information access and learning. The next largest category, work and productivity, trails far behind at 14%. For now, most people experience AI as something that makes them better informed, not something that threatens their livelihood.
Has AI made daily life better or worse?
"So far, what has been the overall impact of AI on your daily life?" — (n=1,042)
FIGURE 4
What has AI changed most in people’s lives?
Coded categories from open-ended responses on the most noticeable changes from AI — (n=1,045)
FIGURE 5
Your economic situation shapes your AI attitudes
Overall, respondents rate the quality of life improvement over the past 50 years at 3.63 out of 5, and their expectations for the next 50 years at 3.67 (Charts 6, 7). Each step up the income ladder corresponds to roughly 0.3 points more optimism on both past and future assessments, from 3.00/3.16 for struggling respondents to 3.97/3.90 for comfortable ones.
Does income shape how people assess the past and future?
"Life is better than 50 years ago" vs. "Life will be better in 50 years"
FIGURE 6
At the regional level, Sub-Saharan Africa and South Asia are the only regions where future expectations significantly exceed assessments of past progress (4.39 vs 3.97, and 4.02 vs 3.71 respectively). Western Europe, East Asia and Oceania tilt the other way.
The starkest difference is Eastern Europe, where respondents give the past 50 years the highest score of any region (4.06) but rate the next 50 years among the lowest (3.22). People in these countries have experienced transformative improvement since the end of the Cold War but appear sceptical that the coming decades will deliver anything comparable.
Does region shape how people assess the past and future?
"Life is better than 50 years ago" vs. "Life will be better in 50 years"
FIGURE 7
Overall
1,041
Western Europe
139
Sub-Saharan Africa
144
South Asia
227
Southeast Asia
82
Middle East & N. Africa
80
Eastern Europe
53
East Asia
104
Latin America
89
North America
136
Oceania
13
1
2
3
4
5
Very Pessimistic
Very Optimistic
Quality of life compared to 50 years ago
Expectations of quality of life 50 years from now
People's existing material circumstances consistently shape how they see AI's future. Financially comfortable respondents are far more likely to believe AI benefits will reach them personally, at 59%, compared to 40% among those who are struggling (Chart 13). The gradient is steady across all four income groups, suggesting this is not simply an optimism gap between the very rich and very poor.
Does financial comfort predict confidence in AI’s benefits?
"On our current trajectory, how likely are the economic benefits of AI to reach you personally?" - by economic situation
FIGURE 8
Wealthier regions show a similar pattern in how they'd prefer to benefit from AI-driven productivity gains. Western European respondents prefer less work over more money by a 56-33% margin, while South Asian and Sub-Saharan African respondents tilt heavily toward more money (62% and 61% respectively; Chart 12). Globally the split is close to even (41% vs 48%), but the regional variation is wide. This likely reflects basic material priorities; where incomes are lower and public services thinner, the abstract appeal of leisure gives way to the concrete need for higher earnings.
Would you rather work less or earn more?
"If AI made you more productive, how would you prefer to benefit?" — by region
FIGURE 9
Overall
1,036
21%
20%
11%
28%
20%
Western Europe
139
28%
29%
11%
18%
15%
Sub-Saharan Africa
144
19%
13%
7%
27%
34%
South Asia
227
14%
15%
9%
39%
23%
Southeast Asia
82
20%
18%
5%
35%
22%
Middle East & N. Africa
80
11%
33%
11%
28%
17%
Eastern Europe
53
25%
16%
6%
31%
22%
East Asia
104
24%
21%
19%
27%
9%
Latin America
89
25%
20%
10%
25%
20%
18%
North America
136
24%
19%
19%
19%
19%
Oceania
13
31%
31%
23%
8%
8%
Strongly prefer less work
Somewhat prefer less work
No preference
Somewhat prefer more money
Strongly prefer more money
Survey participants were presented with two visions of a future where AI handles most tasks:
In World A, the "Guaranteed Jobs Society," human roles are maintained in areas like teaching, caregiving and skilled craftsmanship, with everyone guaranteed a job under shorter hours and better conditions.
In World B, the "Guaranteed Income Society," jobs are no longer central to life. Everyone receives a guaranteed income and people are free to study, create, volunteer or spend time with family, with social recognition tied to community contribution rather than employment.
Overall, more people want guaranteed jobs (52%) than guaranteed income (39%), with 9% expressing no preference (Chart 20).
Guaranteed jobs or guaranteed income — which future do people prefer?
"Which world would you rather live in?" (n=1,024)
FIGURE 10
Guaranteed Jobs Society
Work stays central. Government guarantee jobs for all.
Guaranteed Income Society
AI does most work. Everyone receives enough to live well.
31%
21%
9%
20%
19%
Strongly:
Guaranteed Jobs
Somewhat:
Guaranteed Jobs
No
preference
Somewhat:
Guaranteed Income
Strongly:
Guaranteed Income
The guaranteed jobs vs. guaranteed income question also reveals how much current employment shapes preferences about the future. Among those with good jobs, 56% prefer World A and only 36% choose World B. Those without good jobs flip, with 51% preferring World B and 40% choosing World A. The middle group, those only somewhat satisfied, stays close to the good-job group at 54%. This suggests the preference shift toward guaranteed income is concentrated among people who are genuinely dissatisfied with their work, rather than those who are merely ambivalent.
Does job satisfaction shape which future people want?
Guaranteed Jobs vs. Guaranteed Income preference — by current job satisfaction
Pessimists about their family's future are more likely to frame wealth-sharing as compensation for job displacement (43%) than optimists are (29%). Optimists lean instead toward the idea that AI is built on shared knowledge and its benefits should therefore be shared (50% vs 37% among pessimists; Chart 43f). The reasoning people reach for reflects the future they expect, with those who see AI as a threat anchoring their case in loss. Those who see it as an opportunity appeal to collective ownership. This distinction may matter for policy design, since compensation framing implies a temporary remedy while shared-knowledge framing implies a permanent entitlement.
Do optimists and pessimists justify wealth-sharing differently?
Rows: "How likely is it that life for your family will be profoundly better in the next 50 years?"
Segments: "Which best explains why people should receive a share of the wealth created by AI?"
FIGURE 12
People want work and economic security
A person's economic situation is also a strong predictor of whether they feel they have a good job. Half of financially comfortable respondents say they have a good job by their own definition, compared to just 12% of those who are struggling (Chart 21c). Nearly half of struggling respondents say their job is not good, while a small majority of the “struggling” and “getting by” groups are ambivalent.
Who gets to say they have a “good job”?
"Based on what matters to you, do you have a “good” job?" — by household economic situation
FIGURE 43
Overall
1,024
32%
52%
16%
Struggling
43
12%
42%
46%
Stretched
222
19%
53%
28%
Getting by
467
25%
59%
16%
Comfortable
292
50%
42%
8%
Good job
Okay
Not good
But what is it that makes a job “good”? When asked, respondents prioritize material stability over meaning, dignity, and even safety. Good pay, reasonable hours and time for family, and job security cluster tightly at the top. Meaningful work that helps others comes fourth at 16%, with respect, dignity and safe working conditions well behind at 5% each (Chart 17).
What makes a job "good"?
Percentage selecting each factor in their top two priorities (n=1,026)
FIGURE 13
Good pay
21%
Reasonable hours / time for family
21%
Job security and stability
20%
Meaningful work that helps others
16%
Opportunity to learn and advance
12%
Respect and dignity
5%
Safe working conditions
5%
0
5
10
15
20
25
But this hierarchy shifts depending on people's current circumstances. Among those without good jobs, 35% rank good pay as the most important factor. Among those who already have good jobs, pay drops to 12% and meaningful work rises to 20%, becoming their top priority (Chart 18). This is a straightforward Maslow effect: where basic needs are met, people value time; where they aren't, they value money. It corroborates the pattern shown in chart 12, where richer regions are more likely to favor free time over additional income.
Do people with good jobs want different things than those without?
"What makes a job ‘good’?" — by whether respondents currently have a good job
FIGURE 14
Good pay
35%
22%
12%
Reasonable hours / time for family
18%
23%
18%
Job security and stability
23%
20%
18%
Meaningful work that helps others
12%
14%
20%
Opportunities to learn and advance
7%
12%
15%
Respect and dignity
3%
4%
9%
Safe working conditions
3%
4%
8%
0
5
10
15
20
25
30
35
40
Yes, my job is a "good" job (272)
Somewhat, my job is okay (449)
No, my job is not a "good" job (146)
People don’t want their jobs to be automated
Forty percent of respondents believe their own job is likely to be automated within ten years, but only 21% think it should be (Chart 36). Opinion on whether automation is coming is genuinely split, with 40% saying likely, 39% unlikely and 20% unsure. But on whether it should happen, there is no such ambiguity: 65% say no. People are uncertain about AI's trajectory but are confident that they would prefer for it not to take their job.
Two thirds also say their job makes a meaningful contribution to the world, which may help explain the reluctance. People are not simply protecting their income. Many believe the work itself matters, and that automating it away would be a loss regardless of whether it is technically feasible.
Do people think their jobs are meaningful — and should they be automated?
Three questions on job meaning, automation likelihood, and whether automation should happen (n=1,042)
FIGURE 15
Even among people who see automation coming, fewer than half (45%) endorse it. 43% say it shouldn't, even though they believe it will. Among those who think automation is unlikely, resistance is near-universal at 91%.
Do people who expect automation also want it?
"Do you think your job should be automated?" — grouped by expectation of automation
FIGURE 16
This resistance to automation is not irrational. When asked to pick their top three priorities for the future, 35% of respondents select “meaningful work”, placing it third behind only healthcare (43%) and food and water (38%; Chart 8), and above safety. Work therefore ranks alongside basic survival needs rather as a luxury.
Yet current satisfaction with meaningful work scores just 2.78 out of 5, among the lowest of any domain and below the scale midpoint (a value of 3 represents “adequately met”). People place enormous value on meaningful work while reporting that they don't have enough of it.
Combined with the two thirds who say their current job contributes meaningfully to the world (Chart 36) and the 52% who prefer a future of guaranteed good jobs over guaranteed income (Chart 20), the picture is consistent: people want work that matters. Many feel they already have it, and most of the rest want more of it, not less.They value work as a source of purpose, and they want AI to enhance that rather than replace it.
Are the things people value most also the things they have?
Priority ranking vs. current satisfaction across life domains — overall (n=1,037)
FIGURE 17
Healthcare
43%
3.3
Food & water
38%
3.3
Meaningful work
35%
2.8
Safety
33%
3.0
Housing & infrastructure
33%
3.2
Environment
30%
3.0
Governance
24%
2.6
Education
22%
3.4
Community
6%
2.9
Social support
5%
2.9
0
10
20
30
40
50
% selecting as priority
Current satisfaction
Priority gap: Meaningful work is both highly prioritized AND poorly provided.
When asked how AI will affect specific areas of life over the next decade (Chart 33), one domain is a clear outlier. 60% of respondents expect AI to reduce the availability of good jobs, making it the only area where negative expectations clearly dominate. The outlook is more favourable across other domains. 69% percent expect AI to improve their free time, and slim majorities see benefits for cost of living (51%) and community well-being (53%). The question of whether AI will affect people's sense of purpose is less settled, with the largest group (39%) expecting no major change and opinion leaning slightly positive among the rest (37% better vs 24% worse).
The contrast with free time is worth noting. People are confident AI will give them more time, but less sure it will give them something meaningful to do with it. That tension between freed-up time and uncertain purpose may be one of the more important social questions AI raises.
Where will AI help — and where could it hurt?
"Do you think the increased use of AI across society is likely to make this better, worse or stay the same?" (n=1,042)
FIGURE 18
Cost of living
7%
44%
23%
21%
6%
Free time
15%
54%
20%
8%
3%
Community well-being
9%
44%
26%
17%
4%
Availability of good jobs
5%
20%
14%
40%
20%
Sense of purpose
7%
30%
39%
18%
6%
Profoundly better
Noticeably better
No major change
Noticeably worse
Profoundly worse
When rating the risks and benefits of specific technologies (Chart 28), respondents draw a clear line based on human control. Messaging apps (56% net positive) and AI chatbots (51% net positive) are seen favourably, technologies where people remain in charge of the interaction. Social media sits in the middle, split almost evenly between positive and negative assessments. But AI performing tasks without supervision (50% say risks outweigh benefits) and AI outperforming humans on most valuable work (49%) are the only two technologies where a plurality sees net harm. The pattern is consistent with everything else in the survey: people welcome AI that assists them and resist AI that replaces them.
Which technologies do people see as beneficial — and which as risky?
"Considering both potential benefits and risks, how do you assess the overall impact on society?" (n=1,045)
FIGURE 19
Messaging apps
24%
32%
28%
12%
4%
Social media apps
10%
24%
29%
24%
13%
AI chatbots
18%
33%
27%
15%
7%
AI performing tasks without supervision
8%
18%
24%
28%
22%
AI outperforming humans on most valuable work
10%
19%
22%
23%
25%
Benefits far outweigh risks
Benefits slightly outweigh risks
Equal
Risks slightly outweigh benefits
Risks far outweigh benefits
People want AI to fill gaps, not replace roles
If people don't want AI taking jobs, where do they want it? The answer is strikingly consistent. When asked what would make AI beneficial for their community, respondents converge on the same three themes: creating and protecting jobs, making healthcare accessible, and improving education quality (Chart 16). This trio appears in the top three for seven of the ten regions surveyed, with the order varying but the priorities remaining stable. These priorities also track closely with what people say matters most for their own wellbeing: healthcare (43%), food and water (38%), and meaningful work (35%; Chart 8).
What would make AI beneficial for your community?
Coded themes from open-ended responses on what would make AI beneficial (n=1,030)
FIGURE 20
Create/protect jobs
21%
Make healthcare accessible
21%
Improve education quality
21%
Be accessible to all
16%
Be affordable/free
12%
Be controlled locally
10%
Improve government/public services
8%
Fight corruption
7%
Provide clean energy/water
6%
Ensure safety/privacy
6%
Reduce inequality
5%
Improve infrastructure
5%
Help with farming
3%
Work in local languages
1%
0
5
10
15
20
25
Struggling respondents report much lower satisfaction across every domain than comfortable ones, with gaps as wide as 0.87 points on meaningful work and 0.82 on housing (Chart 5). But their expectations of AI narrow that gap considerably (Chart 6). On healthcare, for example, struggling respondents rate current satisfaction at 2.80 but expect AI impact at 3.60, a jump of 0.80 points. Comfortable respondents start higher (3.47) but expect a smaller uplift (to 3.97). The pattern holds across most domains. Those with the least are not the most cynical about AI; they are less optimistic in absolute terms, but they see more room for improvement and expect AI to deliver some of it. The exception is community, where struggling respondents rate AI's expected impact (2.69) below their current satisfaction (2.78), the only domain for any income group where people expect AI to make things actively worse.
How well are needs met across income levels?
Current satisfaction by household income situation — mean scores on 5-point scale
FIGURE 21
Healthcare
Education
Food & water
Housing &
infrastructure
Safety
Governance
Social
support
Meaningful
work
Environment
Community
Leisure time
Struggling
45
2.8
3.0
2.9
2.4
2.6
2.1
2.4
2.2
2.4
2.8
2.6
Stretched
226
3.1
3.2
3.2
2.9
2.9
2.4
2.6
2.5
2.6
2.8
3.0
Getting by
474
3.3
3.4
3.3
3.1
3.0
2.6
2.9
2.8
2.8
2.9
3.2
Comfortable
296
3.5
3.5
3.4
3.2
3.1
2.8
3.1
3.1
3.0
3.2
3.4
Not met
Fully met
Respondents are most optimistic about AI’s impact on healthcare, education, and leisure time, and most pessimistic about its effects on the environment, community bonds, governance, and employment (Chart 6). The top and bottom of this ranking are revealing. The domains where people expect AI to help most are those that can be improved through better information and efficiency: diagnosing illness, personalising learning, automating routine tasks. The domains where expectations are lowest involve human relationships, cooperation, collective decision-making and a sense of purpose, things that are harder to scale or automate.
How do people expect AI to change things?
Expected AI impact by household economic situation — mean scores (1=much worse, 3=same, 5=much better)
FIGURE 22
Healthcare
Education
Food & water
Housing &
infrastructure
Safety
Governance
Social
support
Meaningful
work
Environment
Community
Leisure time
Struggling
45
3.6
3.8
3.1
3.0
3.2
2.6
3.1
2.8
2.6
2.7
3.2
Stretched
226
3.8
3.9
3.2
3.4
3.3
2.9
3.2
3.0
2.9
2.7
3.6
Getting by
474
3.9
3.9
3.3
3.4
3.3
3.0
3.2
3.0
3.0
2.8
3.6
Comfortable
296
4.0
4.0
3.4
3.5
3.4
3.1
3.4
3.2
3.1
3.1
3.7
Not met
Fully met
The hope matrix: where can AI help most?
Current satisfaction vs. expected AI impact
MATRIX CHART PENDING
Survey participants were asked: If AI automates many jobs and people spend less time working, which would you most want for your community? 57% choose free public services, far ahead of lower prices (23%), direct cash (10%), or AI tools for communities (11%; Chart 10). The preference for public services holds across every income group and every region in the survey, with remarkably little variation across income brackets (53-59%). This is a notable finding given ongoing policy debates around universal basic income. When offered the choice, people across very different economic circumstances prefer collective provision over individual transfers.
If AI automates jobs, what do you want the most?
"If AI automates many jobs and people spend less time working, which would you most want for your community?" - by household economic situation
FIGURE 27B
Free public services
Overall
1,041
57%
Struggling
45
53%
Stretched
226
54%
Getting by
474
57%
Comfortable
296
59%
0
10
20
30
40
50
60
70
FIGURE 27C
Lower prices on goods
Overall
1,041
23%
Struggling
45
24%
Stretched
226
27%
Getting by
474
22%
Comfortable
296
20%
0
10
20
30
40
50
60
70
FIGURE 27D
Cash distributed directly
Overall
1,041
10%
Struggling
45
11%
Stretched
226
12%
Getting by
474
9%
Comfortable
296
9%
0
10
20
30
40
50
60
70
FIGURE 27E
AI tools for communities
Overall
1,041
11%
Struggling
45
11%
Stretched
226
7%
Getting by
474
11%
Comfortable
296
13%
0
10
20
30
40
50
60
70
The regional picture is broadly similar, though direct cash payments find more support in Eastern Europe (22%) and North America (21%) than in Southeast Asia (2%) or Sub-Saharan Africa (3%), where public service provision may feel like a more pressing gap to fill. Western Europe is one of the lowest for free public services but highest on lower prices on goods. Presumably since many of these countries already have relatively affordable public services but are currently battling cost-of-living crises
Sub-Saharan African respondents are the most enthusiastic about AI tools for communities (19%, tied with their preference for lower prices), suggesting that where public service infrastructure is weakest, people see more value in direct tool provision rather than expanded government programs.
That Sub-Saharan Africa and Southeast Asia are the least interested in cash transfers is perhaps the most counterintuitive result here. It may reflect a view that direct payments, however welcome, do not build anything lasting. The provision of AI tools for communities can help find local solutions to pressing challenges. Public services create local infrastructure, local jobs and local expertise. Cash does not. People in these regions may be less interested in receiving AI's dividends than in having AI build and strengthen the systems they depend on.
If AI automates jobs, what do you want the most?
"If AI automates many jobs and people spend less time working, which would you most want for your community?" - by region
FIGURE 28A
Free public services
Overall
1,036
57%
Western Europe
137
47%
Sub-Saharan Africa
142
59%
South Asia
224
63%
Southeast Asia
82
65%
Middle East & N. Africa
75
61%
Eastern Europe
51
51%
East Asia
100
59%
Latin America
88
55%
North America
124
45%
Oceania
13
62%
0
10
20
30
40
50
60
70
FIGURE 28B
Lower prices on goods
Overall
1,036
23%
Western Europe
137
33%
Sub-Saharan Africa
142
19%
South Asia
224
20%
Southeast Asia
82
24%
Middle East & N. Africa
75
21%
Eastern Europe
51
20%
East Asia
100
19%
Latin America
88
25%
North America
124
24%
Oceania
13
15%
0
10
20
30
40
50
60
70
FIGURE 28C
Cash distributed directly
Overall
1,036
10%
Western Europe
137
13%
Sub-Saharan Africa
142
3%
South Asia
224
4%
Southeast Asia
82
2%
Middle East & N. Africa
75
9%
Eastern Europe
51
22%
East Asia
100
13%
Latin America
88
12%
North America
124
21%
Oceania
13
15%
0
10
20
30
40
50
60
70
FIGURE 28D
AI tools for communities
Overall
1,036
11%
Western Europe
137
7%
Sub-Saharan Africa
142
19%
South Asia
224
13%
Southeast Asia
82
9%
Middle East & N. Africa
75
8%
Eastern Europe
51
8%
East Asia
100
9%
Latin America
88
8%
North America
124
10%
Oceania
13
8%
0
10
20
30
40
50
60
70
Some regions are more optimistic than others that AI will work there
Regional variation in AI optimism is large and does not follow the pattern you might expect. When asked whether AI benefits will reach their region, Sub-Saharan African respondents are by far the most confident, with 80% saying it's likely (Chart 13b). Southeast Asians, South Asians, and North Americans form a middle tier. The most skeptical are Eastern Europeans, Latin Americans, and East Asians. The pattern on whether wealthy-country AI is useful locally (Chart 14) largely mirrors the confidence question. North Americans and Sub-Saharan Africans are the most positive. East Asians and Latin Americans are the most skeptical.
Which regions are most confident AI’s benefits will reach them?
"On our current trajectory, how likely are the economic benefits of AI to reach you personally?"
FIGURE 29
Overall
1,082
24%
24%
52%
Western Europe
139
33%
29%
38%
Sub-Saharan Africa
144
10%
10%
80%
South Asia
227
18%
27%
55%
Southeast Asia
82
16%
23%
61%
Middle East & N. Africa
80
25
31%
44%
Eastern Europe
53
43%
25%
31%
East Asia
104
21%
38%
41%
Latin America
104
43%
10%
47%
North America
136
25%
24%
51%
Oceania
13
42%
17%
42%
Unlikely
Neutral
Likely
The pattern on whether wealthy-country AI is useful locally (Chart 14) largely mirrors the previous chart. North Americans and Sub-Saharan Africans are the most positive. East Asians and Latin Americans are the most skeptical.
Will AI built in wealthy countries be useful elsewhere?
"Do you think AI systems developed in wealthy countries will be useful for solving problems in your community?"
FIGURE 30
Overall
1,082
12%
46%
19%
16%
7%
Western Europe
139
6%
49%
22%
19%
4%
Sub-Saharan Africa
144
27%
42%
11%
13%
7%
South Asia
227
11%
52%
14%
17%
6%
Southeast Asia
82
12%
44%
26%
12%
6%
Middle East & N. Africa
80
11%
41%
28%
12%
8%
Eastern Europe
53
6%
44%
12%
18%
20%
East Asia
104
8%
38%
32%
19%
3%
Latin America
104
12%
35%
14%
24%
15%
18%
North America
136
11%
54%
17%
19%
5%
Oceania
13
0%
58%
25%
8%
8%
Extremely likely
Somewhat likely
Neither
Somewhat unlikely
Extremely unlikely
Sub-Saharan Africa's high optimism across both measures is consistent with their broader AI attitudes: they also show the highest expected AI impact scores across nearly every domain, from healthcare (4.30/5) to education (4.39/5) to leisure (4.07/5; Chart 7).
This is a surprising result on the surface, since one might expect those furthest from AI's centres of development to be the most sceptical. But Chart 15 offers a plausible explanation: respondents identified specific applications they believe will transfer – business and productivity tools, healthcare, and education – as well as specific barriers that won't, such as cultural relevance, local language support, and understanding of local institutions). The optimism appears to be conditional rather than naive; people are distinguishing between AI's technical capabilities, which they see as portable, and its contextual fit, which they recognise as limited.
How do people expect AI to change things?
Expected AI impact by region — mean scores (1=much worse, 3=no change, 5=much better)
FIGURE 31
Healthcare
Education
Food & water
Housing &
infrastructure
Safety
Governance
Social
support
Meaningful
work
Environment
Community
Leisure time
Western Europe
139
3.6
3.6
3.1
3.0
3.0
2.6
3.0
2.8
2.9
2.5
3.5
Sub-Saharan Africa
144
4.3
4.4
4.0
4.2
3.9
3.5
3.8
3.8
3.8
3.4
4.1
South Asia
227
4.0
4.1
3.3
3.7
3.3
3.2
3.5
3.4
2.9
3.0
3.6
Southeast Asia
82
3.9
4.0
3.2
3.3
3.1
2.9
3.2
3.0
2.8
3.0
3.6
Middle East & N. Africa
80
3.8
3.8
3.2
3.2
3.4
2.9
3.0
3.0
2.8
2.3
3.5
Eastern Europe
53
3.6
3.4
2.9
3.0
3.4
2.6
3.0
2.5
2.7
2.4
3.2
East Asia
104
3.8
3.8
3.2
3.2
3.3
2.9
3.0
2.8
3.0
2.9
3.5
Latin America
104
4.0
3.7
3.1
3.4
3.4
2.6
3.3
2.7
2.6
2.8
3.5
North America
136
3.7
3.7
3.2
3.1
3.3
3.0
3.1
2.7
3.1
3.0
3.8
Oceania
13
3.9
3.8
3.3
3.2
3.2
3.3
3.3
3.2
3.0
2.9
3.6
Expect worse
Expect better
Sub-Saharan Africa’s comparatively high optimism towards western AI may also reflect lived experience with technologies that leapfrogged legacy infrastructure rather than diffusing gradually through existing systems. In contexts such as Kenya (which accounts for 89% of respondents in this regional sample) innovations such as mobile banking (notably M-Pesa) and off-grid solar microgrids have delivered tangible, everyday benefits without requiring the fixed-line banking networks or centralized power grids, typical of wealthier economies. This recent history of direct, visible gains from externally developed technologies may plausibly shape expectations that advanced AI systems could generate similarly practical and immediate value.
What do people think will and won't transfer from AI developed in wealthy countries?
Subtitle: Coded themes from open-ended responses on whether wealthy-country AI will help locally (n=1,032; multi-tag per response)
FIGURE 32
What people think will transfer
Business & productivity
17%
Healthcare & diagnostics
12%
Education & learning
12%
Jobs & employment
8%
Safety & security
3%
Scientific research
2%
Agriculture & food
2%
What people think won't transfer
Local context & culture
11%
Governance & corruption
5%
Affordability & inequality
4%
Infrastructure & access gaps
3%
Data bias
2%
There's a gender gap: women are more pessimistic than men about AI's benefits compared to its risks
Whilst both women and men on the whole have generally positive attitudes about AI, men are consistently more optimistic across every measure in the survey (Chart 38). The gap is widest on perceived personal benefit (11 points) and narrowest on direct daily experience (5 points). As questions move from present experience to future expectation, men’s attitudes toward AI become increasingly optimistic relative to women.
Does gender influence attitudes toward AI impacts?
Percentage responding "somewhat likely" or "extremely likely"
FIGURE 33
AI benefits will reach me
58%
47%
+11pp
AI could make better decisions than govt
43%
35%
+8pp
Future quality of life will improve
70%
62%
+8pp
Trust AI chatbot to act in my interest
59%
53%
+6pp
AI has made my daily life better
76%
71%
+5pp
0
10
20
30
40
50
60
70
80
Male (533)
Female (520)
Women also see significantly more risk in advanced AI scenarios. 54% of women say the risks of AI outperforming humans on valuable work outweigh the benefits, compared to 42% of men. Men are split roughly evenly, while women tilt clearly negative (Chart 39).
Does gender influence attitudes toward AI risk?
"How do you assess the overall impact on society of AI systems that can outperform humans on most economically valuable work?"
FIGURE 34
Male
518
14%
21%
23%
19%
23%
Female
511
7%
17%
22%
27%
27%
Benefits far outweigh risks
Benefits slightly outweigh risks
Risks and benefits are equal
Risks slightly outweigh benefits
Risks far outweigh benefits
When asked about expected AI impact on specific areas of life, men are more optimistic than women on every single domain (Chart 41).
Does the gender gap in AI optimism hold across every life domain?
"What impact do you think AI will have on each of these areas?" — % saying "somewhat better" or "much better"
FIGURE 35
Leisure time
55%
65%
+10pp
Meaningful work
36%
44%
+8pp
Education
68%
75%
+7pp
Food & water
39%
46%
+7pp
Safety
42%
49%
+7pp
Governance
28%
35%
+7pp
Healthcare
72%
78%
+6pp
Social support
43%
48%
+5pp
Environment
34%
39%
+5pp
Housing & infrastructure
50%
54%
+4pp
Community
30%
34%
+4pp
0
10
20
30
40
50
60
70
80
Male (515)
Female (508)
Gender gap: what matters most vs. where AI is expected to help
Men are more optimistic about AI impact across every domain, even where women prioritize more
FIGURE 36
← What matters most
41%
46%
38%
38%
37%
33%
31%
35%
36%
30%
32%
28%
23%
32%
23%
22%
24%
23%
7%
6%
5%
5%
40
50
30
20
10
0
% selecting as a top-3 priority
Healthcare
Food & water
Meaningful work
Safety
Leisure time
Housing & infrastructure
Environment
Education
Governance
Community
Social support
Expected AI Impact→
78%
72%
46%
39%
44%
36%
49%
42%
63%
55%
54%
50%
39%
34%
75%
68%
35%
28%
34%
30%
48%
43%
0
10
20
30
40
50
60
70
80
% saying AI will make this "somewhat better" or "much better"
Male (515)
Female (507)
FIGURE 37
← What matters most
41%
46%
38%
38%
37%
33%
31%
35%
36%
30%
32%
28%
23%
32%
23%
22%
24%
23%
7%
6%
5%
5%
40
50
30
20
10
0
% selecting as a top-3 priority
Healthcare
Food & water
Meaningful work
Safety
Leisure time
Housing & infrastructure
Environment
Education
Governance
Community
Social support
Expected AI Impact →
31%
23%
17%
14%
18%
13%
20%
16%
25%
20%
22%
19%
14%
12%
40%
30%
10%
7%
11%
11%
18%
13%
0
10
20
30
40
50
% saying AI will make this "somewhat better" or "much better"
Male (515)
Female (507)
Low trust in government, higher trust in AI
People trust AI chatbots more than they trust their elected representatives. This is a survey of people who use AI regularly, so high chatbot trust is not surprising in itself. What is more striking is how low institutional trust is by comparison, and how naturally AI slots into the space between trusted personal advisors and distrusted institutions. For policymakers hoping to shape how AI is used, this is likely to be an uncomfortable finding. People are more likely to take guidance from a chatbot than from the government trying to regulate it.
Who do people trust to act in their best interest?
"To what extent, if at all, do you trust [entity] to act in your best interest? (n=1,044)
FIGURE 38
Your family doctor
42%
43%
10%
4%
1%
Your social media feed (e.g. TikTok, Facebook)
3%
14%
25%
31%
26%
Your elected representatives
3%
28%
21%
28%
20%
Your faith or community leader
11%
34%
29%
15%
10%
Civil servants in your government
4%
28%
25%
27%
16%
Your AI chatbot (e.g. ChatGPT)
12%
44%
28%
12%
4%
Strongly trust
Somewhat trust
Neither
Somewhat distrust
Strongly distrust
At the institutional level, public research institutions command the most trust (70%), followed by small businesses (52%) and public utility companies (43%). As with the previous chart, governments are considered by respondents to be one of the least trusted institutions, second only to social media companies. AI companies are trusted more than big tech and government for now, but they sit closer to the distrusted end of the spectrum.
Which institutions do people trust to do what’s right?
"To what extent, if at all, do you trust [institution] to do what is right? (n=1,044)
FIGURE 39
Governments
3%
24%
17%
32%
24%
Small businesses
8%
44%
28%
16%
4%
Large corporations
6%
22%
21%
28%
23%
Social media companies
2%
13%
22%
30%
32%
Companies building AI
6%
27%
27%
24%
16%
Public utility companies
5%
38%
32%
18%
7%
Public research institutions
19%
51%
20%
8%
2%
Strongly trust
Somewhat trust
Neither
Somewhat distrust
Strongly distrust
This level of distrust in government is grounded in experience. As we saw in chart 5, when asked to rate how well their needs are being met across eleven life domains, respondents rank governance last at 2.6 out of 5, below even meaningful work and environment.
When asked directly whether AI or elected governments should make decisions about how AI affects people's lives, the public is split: 38% agree that AI would do better than governments, 28% disagree, and 34% are unsure (Chart 32). This is not a ringing endorsement of being governed by AI systems, but given the low trust in governments, the fact that more people agree than disagree is notable.
Could AI make better decisions than elected governments?
Agreement with: "AI could make better decisions on my behalf than my government representatives" (n=1,042)
FIGURE 40
AI wealth should be shared because it’s built on global knowledge.
Nearly half of respondents (47%) say the primary reason to share AI-generated wealth globally is that AI is built on shared human knowledge, and its benefits should therefore be shared. A further 33% frame it as compensation for job displacement (Chart 43).
The dominance of the shared knowledge argument is notable, as it frames wealth-sharing as a matter of principle rather than remedy, something people are owed because of what AI is built on, not because of what it might take away. The compensation framing, by contrast, is conditional on harm. That nearly half the sample reaches for the stronger, unconditional argument suggests broad public appetite for treating AI wealth distribution as a right rather than a safety net.
Why do people believe AI wealth should be shared?
"Which best explains why people should receive a share of the wealth created by AI?" (n=1,068)
FIGURE 41
The "shared knowledge" argument is also the plurality choice in almost every region, income group, and demographic slice, with one clear exception: the Middle East and North Africa, where job displacement compensation leads at 50% (Chart 43b). This may reflect the particular intensity of employment anxiety in the region. MENA respondents show the strongest preference of any region for guaranteed jobs over guaranteed income (61% vs. 28%; Chart 20b), and youth unemployment across the Middle East and North Africa has historically ranked among the highest in the world. In a context where secure employment is already scarce, the prospect of AI displacing jobs further may feel less like a theoretical risk and more like an acceleration of an existing crisis.
Does the justification for sharing AI wealth vary by region?
"Which best explains why people should receive a share of the wealth created by AI?" - by region
FIGURE 42A
Built on shared knowledge, so benefits should be shared
Overall
1,015
47%
Western Europe
136
38%
Sub-Saharan Africa
142
55%
South Asia
219
48%
Southeast Asia
81
47%
Middle East & N. Africa
74
28%
Eastern Europe
48
42%
East Asia
96
54%
Latin America
83
49%
North America
124
49%
Oceania
12
42%
0
10
20
30
40
50
60
FIGURE 42B
Compensation for job losses from automation
Overall
1,015
33%
Western Europe
136
43%
Sub-Saharan Africa
142
28%
South Asia
219
30%
Southeast Asia
81
31%
Middle East & N. Africa
74
50%
Eastern Europe
48
35%
East Asia
96
23%
Latin America
83
31%
North America
124
30%
Oceania
12
33%
0
10
20
30
40
50
60
FIGURE 42C
Payment for personal data used to train AI
Overall
1,015
13%
Western Europe
136
9%
Sub-Saharan Africa
142
11%
South Asia
219
16%
Southeast Asia
81
15%
Middle East & N. Africa
74
16%
Eastern Europe
48
12%
East Asia
96
14%
Latin America
83
13%
North America
124
11%
Oceania
12
8%
0
10
20
30
40
50
60
FIGURE 42D
Maintain economic stability amid disruption
Overall
1,015
8%
Western Europe
136
11%
Sub-Saharan Africa
142
6%
South Asia
219
6%
Southeast Asia
81
7%
Middle East & N. Africa
74
5%
Eastern Europe
48
10%
East Asia
96
9%
Latin America
83
6%
North America
124
10%
Oceania
12
17%
0
10
20
30
40
50
60