COUNTRIES WILL BE MOST AFFECTED BY AI AUTOMATION

A WORLD MAP OF THE INVASION OF ARTIFICIAL INTELLIGENCE TAKING OVER JOBS

For many members of the global workforce, the rise of artificial intelligence spells doom. The majority of the employed population feels it is only a matter of time before human-like computer intelligence takes over the workforce. But is there any truth to these fears? Will AI take over jobs, and if so, which labor groups and countries will be most affected by AI automation? We have compiled this report to answer such questions about the impact of AI on the job market.

HOW DID WE DO?

First, we get data from WillRobotsTakeMyJob.com on the automation risk of each job. According to Willrobotstakemyjob, the automation risk is calculated based on the abilities, knowledge, skills and activities required in order to do the job. The author used the same method as a 2013 study by Professor Carl Benedikt Frey and Professor Michael A. Osborne from Oxford Martin School. You can find more details on how the site calculates the automation risk here. Second, we put those jobs into professional groups to find out what the average risk of automation is for each group. To figure out how much of each country's workforce is highly exposed to AI and robot automation, we use data from the International Labor Organization (ilo.org) about how many people work in each occupational group in each country. We then combine this information with the average automation risk of each occupational group. The methodology is described in detail at the end of this article.

   

KEY FINDINGS

64%

of [worldwide workforce] is at high risk of AI automation.

74%

of [US workforce] is at high risk of AI taking over their jobs in the future, which equals to more than 110 million workers.

key-findings

TOP 3 NATIONS THAT WILL BE HARDEST HIT BY AI JOB AUTOMATION IN EACH REGION

  • AFRICA
    • 1. Zambia
    • 2. Angola
    • 3. Uganda
  • ASIA & PACIFIC
    • 1. Bhutan
    • 2. Pakistan
    • 3. India
  • EUROPE
    • 1. Armenia
    • 2. North Macedonia
    • 3. Estonia
  • MIDDLE EAST
    • 1. Iran
    • 2. Jordan
    • 3. Egypt
  • SOUTH/ LATIN
    AMERICA
    • 1. Bolivia
    • 2. Mexico
    • 3. Saint Lucia

TOP 5 NATIONS MOST / LEAST AFFECTED BY AI AUTOMATION

MOST LEAST
  • zambia Zambia
  • zambia Bhutan
  • zambia Angola
  • zambia Armenia
  • zambia Pakistan
  • zambia Trinidad & Tobago
  • zambia Botswana
  • zambia Slovakia
  • zambia Panama
  • zambia Singapore

JOBS EXPOSING TO HIGHEST RISKS OF AI AUTOMATION:

IndustryOccupationAverage AI Automation Risk
Administrative &
support service activities
Clerical support workers89.5%
Administrative &
support service activities
Service and sales workers85.7%
ManufacturingPlant and machine operators,
and assemblers
86.5%
ManufacturingCraft and related trades workers84.4%
Transportation and storageElementary occupations81.5%
Section 1

A WORLD MAP OF THE AI JOB AUTOMATION INVASION

Every country in the world stands at risk of having its workforce population replaced by AI automation. However, some countries will be more affected than others. This section analyzes the extent of the AI automation invasion per country by calculating the percentage of a country’s labor force that is at high risk of AI automation in the future.

64% Of the world’s Labor Force Is Highly Vulnerable to AI Automation in the Near Future.

More specifically, the World Economic Forum’s Future of Jobs Report 2020 indicated that AI might render at least 85 million of the world’s labor force unemployed by 2020. However, other stakeholders, like Goldman Sachs place the statistics on AI-related job losses at 300 million.

Conversely, the World Economic Forum report also says that AI technology can create at least 97 million new jobs within the same period. So, all is not lost, but there is a need to scrutinize the introduction of AI into the workplace to ensure the technology’s demerits do not outweigh its benefits.

Half or More of the Workforce in 82 of the 85 Countries Studied Is at High Risk of Becoming Automated in The Future.

From our findings, Singapore is the only country with less than a 40% chance of its employed population being gravely affected by the rise of AI work automation. In fact, only 3 of the 85 countries listed in the table above have less than 50% of their workforce heavily threatened by AI.

All of the Top 10 Countries Where Jobs Are Least Likely to Be Lost to Automation Belong to either the High Income or Upper Middle Income Groups.

Countries with high and upper-middle-income rates are less likely to feel the impact of AI job automation.

Africa is the Most Vulnerable to High Risk of Jobs Being Taken by AI.

Africa is a tech hub on the rise. However, most countries on this continent have not embraced work automation extensively, leaving many employees at high risk of losing their jobs once there’s a move to AI.

Percentage of Workforce at High Risk of AI Automation, World Map

A M PK B T SG SK CL T O IE CH L U RS SE HU BG TR IL IQ AF K G KM MU BR DO SV GR C Y JO AE EG IR XK MK BA M T IT FR ES PT HR A T DE NL BE IS TL VN BN KR TH IN RU MN BO AR MX US BZ CR L C EC C O MD RO EE FI NO L T PL L V ZW R W E T PH SI T T UY B W B Y CZ DK P A UG A O ZM
  • 0 30
  • 40
  • 50
  • 60
  • 70
  • 80
  • 100

And that’s our summary of the tabulated findings. Please find the full data in the the table below:

PERCENTAGE OF WORKFORCE AT HIGH RISK
OF AI AUTOMATION, BY COUNTRY

CountryRegionIncome% of workforce at high risk of automation
ZambiaAfricaLower middle income82.76%
BhutanAsia & PacificLower middle income82.48%
AngolaAfricaLower middle income82.41%
ArmeniaEuropeUpper middle income81.33%
PakistanAsia & PacificLower middle income80.58%
UgandaAfricaLow income80.19%
North MacedoniaEuropeUpper middle income79.89%
ZimbabweAfricaLower middle income79.87%
EstoniaEuropeHigh income79.84%
Republic of MoldovaEuropeUpper middle income79.84%
BoliviaSouth/Latin AmericaLower middle income79.81%
IndiaAsia & PacificLower middle income79.40%
MexicoSouth/Latin AmericaUpper middle income79.40%
Saint LuciaSouth/Latin AmericaUpper middle income78.80%
RwandaAfricaLow income78.77%
RussiaEuropeUpper middle income78.71%
KosovoSouth/Latin AmericaUpper middle income78.69%
Timor-LesteAsia & PacificLower middle income78.68%
Bosnia and HerzegovinaEuropeUpper middle income78.65%
EcuadorSouth/Latin AmericaUpper middle income78.64%
EthiopiaAfricaLow income78.32%
NetherlandsEuropeHigh income78.31%
Viet NamAsia & PacificLower middle income78.24%
CyprusEuropeHigh income77.51%
BelgiumEuropeHigh income77.45%
IranMiddle EastLower middle income77.20%
LithuaniaEuropeHigh income77.14%
BelizeSouth/Latin AmericaLower middle income77.12%
LatviaEuropeHigh income77.04%
RomaniaEuropeUpper middle income76.91%
JordanMiddle EastUpper middle income76.86%
ArgentinaSouth/Latin AmericaUpper middle income76.81%
EgyptMiddle EastLower middle income76.78%
United Arab EmiratesMiddle EastHigh income75.73%
CroatiaEuropeHigh income75.66%
ThailandAsia & PacificUpper middle income75.61%
MaltaEuropeHigh income75.43%
FranceEuropeHigh income75.15%
SpainEuropeHigh income74.96%
FinlandEuropeHigh income74.86%
GreeceEuropeHigh income74.83%
PolandEuropeHigh income74.63%
Republic of KoreaAsia & PacificHigh income74.28%
United StatesNorth AmericaHigh income74.18%
MongoliaAsia & PacificLower middle income73.96%
AustriaEuropeHigh income73.40%
NorwayEuropeHigh income73.33%
ItalyEuropeHigh income72.24%
GermanyEuropeHigh income72.17%
Brunei DarussalamAsia & PacificHigh income71.83%
ColombiaSouth/Latin AmericaUpper middle income71.70%
PortugalEuropeHigh income71.62%
Costa RicaSouth/Latin AmericaUpper middle income70.86%
IcelandEuropeHigh income70.18%
SwitzerlandEuropeHigh income69.69%
SwedenEuropeHigh income69.53%
AfghanistanAsia & PacificLow income69.40%
El SalvadorSouth/Latin AmericaLower middle income69.08%
HungaryEuropeHigh income67.89%
LuxembourgEuropeHigh income67.83%
Dominican RepublicSouth/Latin AmericaUpper middle income67.67%
Occupied Palestinian TerritoryArab States- missing data -66.83%
SloveniaEuropeHigh income66.74%
PhilippinesAsia & PacificLower middle income66.57%
ComorosArab StatesLower middle income66.18%
TongaAsia & PacificUpper middle income66.17%
TürkiyeEuropeUpper middle income65.74%
BrazilSouth/Latin AmericaUpper middle income64.74%
MauritiusAfricaUpper middle income64.50%
SerbiaEuropeUpper middle income63.27%
IrelandEuropeHigh income63.07%
IsraelMiddle EastHigh income62.90%
IraqMiddle EastUpper middle income61.01%
BulgariaEuropeUpper middle income60.76%
KyrgyzstanAsia & PacificLow income60.35%
CzechiaEuropeHigh income58.25%
UruguaySouth/Latin AmericaHigh income57.45%
BelarusEuropeUpper middle income56.73%
ChileSouth/Latin AmericaHigh income56.69%
DenmarkEuropeHigh income55.79%
Trinidad and TobagoSouth/Latin AmericaHigh income54.65%
BotswanaAfricaUpper middle income53.18%
SlovakiaEuropeHigh income46.79%
PanamaSouth/Latin AmericaUpper middle income44.57%
SingaporeAsia & PacificHigh income19.22%

REVEAL MORE

We compiled the automation risk data of each job on Willrobotstakemyjob.com and the number of jobs in each occupational category to determine the nations where AI job automation will have the most significant impact.

license

This image is licensed under the Creative Commons Attribution-Share Alike 4.0 International License - www.creativecommons.org/license/by-sa/4.0t

METHODOLOGY

  1. We compile information regarding the automation risk of each job at https://willrobotstakemyjob.com/. There are a total of 897 jobs.
  2. Then, we put these jobs into groups based on what they were. From there, we have the average automation risk of each occupational group.
  3. Then, we use data from https://ilostat.ilo.org/data/# to figure out how many people work in that group of jobs in each country. All together, there are 105 countries.
  4. Now that we know the average risk of AI automation for each job group and how many people work in that group, we can figure out what percentage of the workforce is at a high risk of AI automation. High risk is defined by willrobotstakemyjob.com as 61 percent or more.
  5. Countries impacted the most by AI automation = Countries with the highest percentage of labor force at high risk of AI automation

OTHER DATA SOURCES:

EMBED

To embed please copy and paste the code below.

Section 2

US MAP OF AI JOB AUTOMATION INVASION

The U.S. is also susceptible to the AI job automation invasion. Let's look at how each state in the United States is dealing with the threat of AI replacement.

74.18% Of the US Workforce Is at Significant Risk of Being Replaced by AI in the Future.

Artificial intelligence efforts in the U.S. pose a significant threat to at least 74.18% of jobs in the country. This rate of job loss may not be immediate, but AI’s rapidly growing capabilities can easily replace human labor in the future.

Colorado, Vermont, New Jersey, Connecticut, and Alaska Are the Top Five States With the Highest Percentage of Their Workforce at Risk of Being Replaced by AI.

Two-thirds of the workforce in Colorado, Vermont, New Jersey, Connecticut, and Alaska are at significant risks of being replaced by human-like computer intelligence in the future. These states all have over a 67% chance of an AI workspace takeover.

At 20.4% and 21.1%, respectively, Kentucky and the District of Columbia Have the Lowest Percentage of Their Workforce at Significant Risk of Being Replaced by AI.

Despite major indicators showing that AI can take over the job market, there is some hope for most states in the U.S. Our data shows that at least 30 states have less than a 50% chance of AI taking over jobs. The states with the lowest percentage among these are Kentucky and the District of Columbia, which have a 20.4% and 21.1% chance of having their workforce replaced by artificial intelligence technology, respectively.

Percentage of Workforce at High Risk of AI Automation, U.S. Map

  • 0 30
  • 40
  • 50
  • 60
  • 70
  • 80
  • 100

PERCENTAGE OF WORKFORCE
AT HIGH RISK OF
AI AUTOMATION, BY STATE

RankState% of workforce at high risk of AI automation
1Colorado68.6%
2Vermont68.5%
3New Jersey67.9%
4Connecticut67.1%
5Alaska67.1%
6Alabama66.9%
7Texas64.7%
8Massachusetts64.5%
9Guam62.0%
10Delaware62.0%
11Virginia60.0%
12Indiana58.3%
13Oklahoma57.5%
14Mississippi55.7%
15Nevada55.7%
16Washington53.0%
17California52.7%
18Missouri51.7%
19Nebraska51.6%
20Pennsylvania51.2%
21Rhode Island51.1%
22South Carolina51.1%
23Florida50.6%
24Ohio50.4%
25Illinois48.6%
26Maryland48.2%
27North Carolina46.8%
28Iowa46.2%
29Arkansas46.2%
30Virgin Islands44.4%
31Arizona44.0%
32Wisconsin42.3%
33New Mexico42.1%
34Idaho41.9%
35Montana41.8%
36Louisiana41.3%
37Oregon40.3%
38Utah40.1%
39Wyoming39.8%
40Maine38.9%
41Hawaii38.4%
42West Virginia37.9%
43Tennessee37.4%
44South Dakota37.0%
45Michigan36.2%
46Georgia34.6%
47Puerto Rico33.3%
48Minnesota30.0%
49New York28.5%
50New Hampshire28.2%
51North Dakota28.2%
52Kansas27.7%
53District of Columbia21.4%
54Kentucky20.4%

REVEAL MORE

We compiled the automation risk data of each job on Willrobotstakemyjob.com and the number of jobs in each occupational category to determine the nations where AI job automation will have the most significant impact.

license

This image is licensed under the Creative Commons Attribution-Share Alike 4.0 International License - www.creativecommons.org/license/by-sa/4.0t

METHODOLOGY

  1. We compile information regarding the automation risk of each job at https://willrobotstakemyjob.com/. There are a total of 897 jobs.
  2. Then, we put these jobs into groups based on what they were. From there, we have the average automation risk of each occupational group.
  3. Then, we use data from https://ilostat.ilo.org/data/# to figure out how many people work in that group of jobs in each country. All together, there are 105 countries.
  4. Now that we know the average risk of AI automation for each job group and how many people work in that group, we can figure out what percentage of the workforce is at a high risk of AI automation. High risk is defined by willrobotstakemyjob.com as 61 percent or more.
  5. Countries impacted the most by AI automation = Countries with the highest percentage of labor force at high risk of AI automation

OTHER DATA SOURCES:

EMBED

To embed please copy and paste the code below.

Section 3

JOBS WITH HIGHEST / LOWEST AUTOMATION RISK

Our research also revealed that certain jobs had a higher risk of being replaced by AI than others. By breaking down the industry, occupation, and average AI automation risk, this table and the findings below will help you identify which professions are most endangered.

JOBS WITH HIGHEST / LOWEST AUTOMATION RISK

IndustryOccupation% of workforce at high risk of automation
Administrative &
support service activities
Clerical support workers89.5%
Administrative &
support service activities
Service and sales workers85.7%
ManufacturingPlant and machine operators,
and assemblers
86.5%
ManufacturingCraft and related trades workers84.4%
Transportation and storageElementary occupations81.5%
ManufacturingService and sales workers74.3%
Transportation and storagePlant and machine operators, and assemblers74.6%
Other service activitiesElementary occupations72.3%
ManufacturingProfessionals73.5%
Mining and quarryingPlant and machine operators, and assemblers66.9%
Financial and insurance activitiesProfessionals68.2%
Accommodation and food
service activities
Professionals65.0%
Agriculture; forestry and fishingSkilled agricultural, forestry and fishery workers65.6%
ManufacturingTechnicians and associate professionals64.3%
Other service activitiesService and sales workers57.3%
Administrative &
support service activities
Professionals63.8%
Wholesale and retail trade; repair
of motor vehicles and motorcycles
Service and sales workers55.2%
Other service activitiesTechnicians and associate professionals52.9%
Arts, entertainment and recreationTechnicians and associate professionals52.3%
Transportation and storageProfessionals53.6%
Professional, scientific
and technical activities
Clerical support workers51.0%
Human health and social work activitiesClerical support workers70.8%
Wholesale and retail trade; repair
of motor vehicles and motorcycles
Technicians and associate professionals50.2%
Other service activitiesProfessionals45.8%
Transportation and storageTechnicians and associate professionals60.0%
Human health and social work activitiesTechnicians and associate professionals43.5%
ConstructionProfessionals39.8%
ConstructionTechnicians and associate professionals38.9%
Information and communicationProfessionals43.3%
Information and communicationTechnicians and associate professionals39.5%
Professional, scientific and technical activitiesTechnicians and associate professionals33.9%
Arts, entertainment and recreationProfessionals31.0%
Professional, scientific and technical activitiesProfessionals25.1%
EducationProfessionals22.8%
EducationElementary occupations17.3%
Human health and social work activitiesProfessionals17.6%
Public administration and defence; compulsory social securityProfessionals8.9%

REVEAL MORE

We compiled the automation risk data of each job on Willrobotstakemyjob.com and the number of jobs in each occupational category to determine the nations where AI job automation will have the most significant impact.

license

This image is licensed under the Creative Commons Attribution-Share Alike 4.0 International License - www.creativecommons.org/license/by-sa/4.0t

METHODOLOGY

  1. We compile information regarding the automation risk of each job at https://willrobotstakemyjob.com/. There are a total of 897 jobs.
  2. Then, we put these jobs into groups based on what they were. From there, we have the average automation risk of each occupational group.
  3. Then, we use data from https://ilostat.ilo.org/data/# to figure out how many people work in that group of jobs in each country. All together, there are 105 countries.
  4. Now that we know the average risk of AI automation for each job group and how many people work in that group, we can figure out what percentage of the workforce is at a high risk of AI automation. High risk is defined by willrobotstakemyjob.com as 61 percent or more.
  5. Countries impacted the most by AI automation = Countries with the highest percentage of labor force at high risk of AI automation

OTHER DATA SOURCES:

EMBED

To embed please copy and paste the code below.

Clerical Support Workers in the Field of Administrative and Support Service Activities Face a High Likelihood of Being Replaced by Robots and Artificial Intelligence.

In the near future, clerical workers in the administrative and support service industries have at least a 90% risk of losing their jobs to AI. Trailing closely behind are service and sales workers in the same industry, with an 85.7% chance of job loss.

Next will be plant and machine operators, assemblers, and craft and related trades workers in manufacturing with an 86.5% and 84.4% chance, respectively, and elementary occupations in transport and storage with an 81.5% chance of job loss.

Professionals Employed in the Field of Public Administration and Defense, as Well as Compulsory Social Security, Are Among the Least Likely to Be Affected by the Wave of AI Automation.

The risk of automation-related job loss for employees in public administration, defense, and compulsory social security is as low as 9%. Other occupations with such low stats include elementary education occupations at 17.3%, human health and social work at 17.6%, education professionals at 22.8%, and professional scientific and technical activity workers at 25.1%.

TOP 3 INDUSTRIES WHERE AI AUTOMATION THREATENS EMPLOYMENT THE MOST

81% of manufacturing jobs are at risk of being automated out in the future, followed by 77% of administrative and support service jobs and 70% of transportation and storage jobs.

In contrast, education, with a risk of 18%; human health and social work activities, with a risk of 23%; and public administration and defense; mandatory social security, with a risk of 26%, are the top three industries where jobs are least likely to be replaced by AI and robots in the future.

81.1%
manufacturing
Manufacturing
77.4%
administrative-support
Administrative &
support service
activities
69.7%
transportation-storage
Transportation
& storage
69.7%
agriculture-forestry-fishing
Agriculture,
forestry,
& fishing
61%
quarrying-mining
Mining &
quarrying
69.6%
activities-of-households
Activities of households as
employers; undifferentiated
goods- and services-producing
activities of households for
own use
69.3%
food-service-activities-accommodation
Accommodation &
food service activities
68.1%
insurance-activities-financial
Financial
& insurance activities
58.8%
wholesale-retail-trade
Wholesale & retail
trade; repair of motor
vehicles & motorcycles
54.3%
service-activities-other
Other
service activities
50.2%
electricity-gas-steam
Electricity, gas,
steam, and air
conditioning supply
44.3%
real-estate-activities
Real estate
activities
44.3%
water-supply-sewerage
Water supply; sewerage,
waste management,
& remediation activities
43.5%
construction
Construction
40.3%
communication-information
Information &
communication
38.3%
arts-entertainment-recreation
Arts, entertainment,
& recreation
29.6%
professional-scientific-technical
Professional, scientific,
& technical activities
26.2%
defense-compulsory
Public administration
& defense, compulsory
social security
23.4%
social-work-activities-human-health
Human health &
social work activities
18.0%
education
Education

POSITIONS AT HIGHEST RISK OF BEING REPLACED BY AI

Clerical support positions, plant and machine operators and assemblers, and craft and related trades employees are at the greatest risk of being replaced by AI. Their likelihood of being supplanted by AI in the future exceeds 70%.

OCCUPATION

AVERAGE AUTOMATION RISK (%)

Clerical support workers
78.0% clerical-support-workers
Plant and machine operators,
& assemblers
77.4% plant-and-machine-operators
Craft and related trades workers
73.8% craft-and-related-trades-workers
Service and sales workers
65.6% service-and-sales-workers
Skilled agricultural, forestry,
& fishery workers
51.7% skilled-agricultural-forestry
Elementary occupations
44.3% elementary-occupations
Technicians & associate
professionals
43.7% technicians-associate-professionals
Professionals
35.8% professionals
Managers
28.5% managers
Section 4

DOES WAGE OR EDUCATION IMPACT AUTOMATION RISK OF A JOB?

Further findings reveal that low-education workers are 280% more likely than high-education workers to be at high risk of being replaced by AI. In this case, "higher education" refers to a bachelor's degree or higher educational qualification. Here's a table that breaks down these results.

EDUCATION LEVEL% OF LABOR AT HIGH RISK OF AI AUTOMATION
High Education Labor15%
Low Education Labor43%

Additionally, higher-paying jobs are 34% less likely to be displaced by AI than lower-paying jobs. Low-paying occupations here refer to jobs with wages below the national average wage of $74,738. Below is a table that summarizes these findings.

WAGEAverage AI Automation Risk
High-paying Jobs35.12%
Low-paying Jobs87.81%

METHODOLOGY AND LIMITATIONS

This research followed a structured methodology to achieve the most accurate result possible. Below is a step-by-step explanation of the approach, process, and resources used to obtain the statistics we have presented.

METHODOLOGY

  • We compile information regarding the automation risk of each job at https://willrobotstakemyjob.com/. There are a total of 897 jobs.
  • Then, we put these jobs into groups based on what they were. From there, we have the average automation risk of each occupational group.
  • Then, we use data from https://ilostat.ilo.org/data/# to figure out how many people work in that group of jobs in each country. All together, there are 105 countries.
  • Now that we know the average risk of AI automation for each job group and how many people work in that group, we can figure out what percentage of the workforce is at a high risk of AI automation. High risk is defined by willrobotstakemyjob.com as 61 percent or more.
  • Countries impacted the most by AI automation = Countries with the highest percentage of labor force at high risk of AI automation

OTHER DATA SOURCES:

LIMITATIONS

  • Wage and Education are restricted geographically within the United States.
  • Labor statistics are compiled from the most recent year for which data is available.
  • Other variables may affect the selected metrics, but they were not considered.
  • Incomplete data is omitted.