Global Views On Artificial Intelligence

Introduction

Our project focuses on global perceptions of AI and they may vary across countries and cultures. In 2023 Ipsos conducted a “Global Views on A.I.” report with the aim of better understanding how people across the world feel about artificial intelligence and how they expect it will impact their lives. Major topics in the report included trust in AI, its impact on daily life, and jobs. Later that year, the Pew Research Center and Stanford University's HAI lab published similar reports that referenced the original Ipsos article. Our project takes data from these reports to create unique interactive visualizations that display global perception in AI.

Looking at the Global Scale

This choropleth map uses data from the Pew Research Center 2025 Global attitudes survey. People from all the displayed countries were asked the following question: “How much trust do you have in each of the following countries and organizations to regulate the use of artificial intelligence effectively – a lot of trust, some trust, not too much trust, no trust at all, or are you not sure?” The percentage of people giving each response from each population and for each country were recorded. The choropleth map displays the proportion of people who responded “a lot of trust” or “some trust.” The pages of the map alternate between the country being evaluated. The tooltip is a bar chart that displays the percentage of people who chose each of the categories given in the survey question, excluding "not sure” and "refused," as their sum can be inferred from the other options.

From the choropleth map, we can see that North America and most of Europe have little trust in China’s AI regulation. In contrast, the surveyed African countries, along with Indonesia, have the highest level of trust in China’s AI regulation. The European Union appears to receive the most consistent and least variable level of trust from other countries. In addition, European countries tend to trust the European Union more than the United States. Another interesting thing to note is that the surveyed African countries tend to trust all three countries and organizations presented to them more than most other countries.

How do People View Their Own Country?

The bar chart below displays the proportion of people who responded with “a lot of trust” or “some trust” when asked how much they trust their own country to regulate AI in the Pew survey. Interestingly, the range in the proportion of people who trust their own country to regulate AI is 67%. People in India tend to trust their country the most, while people in Greece trust their country the least.

The Job Market and Your Job

This visualization utilizes data from the HAI lab at Stanford University, presented as an open-source CSV file containing survey responses from employed adults across 32 countries. Economic Freedom Score data was obtained from the Heritage Foundation’s Index of Economic Freedom, with 2024 figures used to align with the 2024 survey data. Countries were then binned into four Economic Freedom Score ranges (Low: 45-55, Mid-low: 55-65, mid-high: 65-75, and High: 75-85) and plotted against two AI optimism metrics: the percentage of respondents expecting AI to improve their specific job and the percentage expecting AI to improve the job market broadly. A global average of all 32 countries is marked as a black diamond for consistent reference across all slides


The visualization reveals a notable inverse trend: countries with lower Economic Freedom Scores tend to cluster toward the upper right of the chart, indicating greater optimism about AI’s impact on both their personal jobs and the job market overall. Higher Economic Freedom Score countries, by contrast, cluster closer to the origin, suggesting more cautious expectations about AI-driven improvements. This pattern may suggest that workers in more state-directed or developing economies view AI as a greater opportunity for upward mobility and job improvement, possibly because they have more to gain from technological disruption. In contrast, workers in highly market-oriented economies may already operate in competitive, dynamic labor markets and therefore hold more cautious or realistic expectations about AI’s net benefit. Though the relationship is completely correlational, it raises an interesting question regarding how the economic systems and labor market structures shape workers’ expectations and attitudes toward emerging technologies like AI.

Will you be replaced by AI?

This graphic utilizes data from the Ipsos article and pew research report. The original data was collected from 14,782 employed adults under the age of 75 across 31 countries. The data was specifically gathered from the HAI Lab at Standford university as an open-source CSV file. Data preprocessing and engineering were conducted to include the total number of people surveyed in addition to the percentage values for each category present in the CSV file. The graph highlights global opinions for the likelihood that AI will change and eventually replace one's job.

Visualizing the Data Through a Different Lens

The first visualization presents an overview of AI perception across the three major regions (Americas, Europe, and Asia-Pacific) with the global average shown as a dashed black reference line. Asia-Pacific extends farthest outward on most axes, indicating the strongest overall optimism, while Europe appears the most contracted, reflecting comparatively lower agreement across the five survey statements. The Americas occupy an intermediate position between the two, demonstrating moderate levels of optimism.

The second visualization disaggregates the Americas into North America and South America, allowing direct comparison between the two sub-regions while keeping the global average visible as a consistent benchmark. North America's profile is heavily driven by the United States, which sits significantly above the global average on most positive dimensions and is clearly carrying the sub-region's overall optimism. Canada and Mexico fall within the global average on every dimension except nervousness, where both countries show elevated responses. In contrast, South America demonstrates strong general optimism across nearly all axes, with Peru standing out as particularly optimistic.

The third visualization breaks down Europe into Western/Northern Europe and Southern/Eastern Europe, again retaining the global average as a reference. Southern/Eastern Europe extends noticeably farther outward than Western/Northern Europe across most axes, indicating relatively greater optimism and agreement within the former group. Western/Northern Europe presents a more cautious and contracted shape, highlighting lower enthusiasm and trust compared to its southern and eastern counterparts. The only axis in which Western/Northern Europe significantly exceeds Southern/Eastern Europe is in their nervousness about AI, further demonstrating their cautious, concerned attitude toward the technology.

Important Note

To make the spiderweb chart readable, each long survey question was shortened into a compact label. These shortened labels are only used for visualization clarity on the axes; all underlying values still come directly from the original survey wording.

Each mapping below shows how the abbreviated label corresponds to the full survey question respondents actually answered:

  • AI will change daily life = Products and services using artificial intelligence will profoundly change my daily life in the next 3–5 years
  • AI is fair (no discrimination) = I trust artificial intelligence to not discriminate or show bias towards any group of people
  • Trust AI data privacy = I trust that companies that use artificial intelligence will protect my personal data
  • Excited about AI = Products and services using artificial intelligence make me excited
  • Nervous about AI = Products and services using artificial intelligence make me nervous

About The Chart

The spiderweb/radar chart visualizations utilize data from the HAI Lab at Stanford University. The data is presented as an open-source CSV file from the same Google Drive folder used for the scatterplot analysis and originally contained survey responses from employed adults across 32 countries. South Africa and Turkey were removed because they did not fit cleanly into the three main geographic regions (Americas, Europe, and Asia-Pacific), leaving 30 countries for the visualizations.

I began with a visualization showing aggregated AI perception trends for the three major regions (Americas, Europe, and Asia-Pacific) as well as the global average (dashed black line). To examine those differences more closely without creating overcrowded charts, I then created two additional spiderweb visualizations that disaggregate two of the regions. The Americas were split into North America and South America, and Europe was split into Western/Northern Europe and Southern/Eastern Europe. Asia-Pacific was not further subdivided because creating clean, non-overlapping, and easily defensible country groupings for that region proved difficult.

Each detailed chart includes all individual countries within its sub-regions, and contains a button to toggle between, for example, both North and South America, just North America, or just South America. The global average is also consistently present in all charts for reference. Overall, this structure first reveals general regional trends and then allows for a focused examination of each sub-region.

Conclusion

An Optimistic Divide between the Developing and Developed

The strongest and most consistent finding across all visualizations is the strong divide in AI optimism between developed and less developed countries. The scatter plot visual demonstrates an inverse relationship between Economic Freedom Score and AI optimism. Countries with lower economies are more optimistic about AI improvements regarding their personal jobs and the market. This is likely due to the possible improvements it may bring to their life. On the other hand, developed countries are much less optimistic, and they fear AI would have a higher chance of taking their jobs and means of living rather than aiding them.

The regional spiderweb visual reinforces this concept. The region that extends the furthest out on the optimism axes is the Asia-Pacific region. While Europe, the most economically developed subregion, is far more contracted and cautious. The choropleth maps also support this, as African countries (a region that is typically considered to be less developed) tend to place more trust in global AI regulation than other regions.

Trust is a Geopolitical Issue

North America and Europe both distrust China's AI regulation and policy, while Africa and Indonesia favor it more heavily. Additionally, European countries trust the EU more than the United States when it comes to AI regulation between the two. This data would suggest that AI regulation and regional opinions are influenced by existing diplomacy and trade relationships.

Excitement and Anxiety

The study we referenced for the data also tracked variables such as "excited about AI" and "nervous about AI" that help capture a much more local opinion for countries surveyed. Many regions score high in both categories, with a great spread between them. For example, Europe is the most nervous and the least excited, while Asia-Pacific is highly excited with lower nervousness. Our spiderweb and choropleth visuals help to back this up.

The BIG Takeaway

Overall, workers in less market-oriented developing countries tend to have more positive views when it comes to AI. They believe this new, accessible technology will pose positive benefits to their day-to-day lives. On the other hand, workers in already competitive developed economies hold more cautious or even skeptical expectations, as many are afraid AI may take their jobs.

So, will AI take your job? Well, probably not in the near future, and it's unlikely even within t he next five years. It's important to stay up-to-date with AI news and informed on new policy and regulations that may affect your country or region.