AI Workforce Displacement: A Briefing for Policymakers

Data-driven evidence on AI job displacement risk, what governments are doing, and policy options backed by research.

US Workforce AI Exposure

BLS verified · Direct assessment 79.4% services · 19% industry · 1.6% agriculture (World Bank / ILO estimates, 2022)

168.7M

workers in zone assessment

0M+

Jobs protected from AI

33% of workers

0M+

Jobs needing adaptation

40% of workers

0M+

Jobs at high risk from AI

27% of workers

US Workforce AI Exposure each figure = ~1 million people

56.2M protected 68.1M transforming 44.3M at risk

Zone Distribution

~168.7M workers

By Workers (BLS employment-weighted)

33% · 56.2M Protected

40% · 68.1M Transforming

27% · 44.3M At Risk

Show distribution by role count (US data)

AI Displacement by State

% workforce in Red Zone

Loading map…

AI Displacement by US State

How many workers in your state face AI displacement? Select a state to see its zone breakdown.

Why does this breakdown look like this?

The US has the highest services-sector share of any major economy (79%). Retail, food service, clerical, and admin roles employ tens of millions of workers — and these routine, rules-based jobs are overwhelmingly classified RED. Meanwhile, healthcare, skilled trades, and education — all requiring physical presence or human judgment — keep roughly 40% of workers in the GREEN zone.

All 51 Ranked by RED Zone %

#	State	Green	Yellow	Red	Workers
1	South Carolina	28.5%	39.0%	32.4%	2.3M
2	Mississippi	31.1%	36.7%	32.2%	1.1M
3	Tennessee	27.9%	40.1%	32.0%	3.3M
4	New Hampshire	29.4%	38.6%	32.0%	666K
5	Indiana	28.3%	39.7%	32.0%	3.2M
6	Georgia	26.7%	41.3%	32.0%	4.8M
7	South Dakota	32.7%	35.5%	31.8%	447K
8	Kentucky	28.4%	39.8%	31.8%	2.0M
9	Michigan	28.3%	40.0%	31.6%	4.4M
10	Alabama	29.8%	38.7%	31.5%	2.1M
11	Utah	28.6%	40.2%	31.3%	1.7M
12	Kansas	31.2%	37.8%	31.0%	1.4M
13	Arkansas	30.1%	38.9%	31.0%	1.3M
14	New Mexico	33.3%	35.8%	31.0%	842K
15	Oklahoma	30.2%	38.8%	31.0%	1.7M
16	North Carolina	29.9%	39.2%	30.9%	4.9M
17	Iowa	29.7%	39.4%	30.8%	1.6M
18	Delaware	29.6%	39.7%	30.7%	461K
19	Nebraska	31.4%	37.9%	30.7%	1.0M
20	Wisconsin	29.7%	39.8%	30.5%	2.9M
21	Idaho	30.1%	39.5%	30.4%	828K
22	New Jersey	30.9%	38.8%	30.3%	4.2M
23	Florida	28.6%	41.1%	30.3%	9.8M
24	Texas	29.9%	40.0%	30.1%	13.7M
25	Arizona	30.3%	39.6%	30.1%	3.2M
26	Missouri	30.9%	39.0%	30.0%	2.9M
27	Ohio	29.7%	40.4%	29.9%	5.5M
28	Washington	32.3%	37.9%	29.8%	3.5M
29	Maine	33.4%	37.0%	29.6%	627K
30	Hawaii	29.4%	41.0%	29.6%	605K
31	Oregon	29.3%	41.2%	29.5%	1.9M
32	North Dakota	34.5%	36.2%	29.3%	417K
33	Vermont	33.7%	37.1%	29.2%	293K
34	Minnesota	32.1%	38.8%	29.1%	2.9M
35	Montana	33.3%	37.6%	29.1%	501K
36	Colorado	30.9%	40.0%	29.1%	2.9M
37	Nevada	27.4%	43.6%	29.0%	1.5M
38	Virginia	31.3%	39.8%	28.8%	4.0M
39	West Virginia	35.9%	35.4%	28.7%	695K
40	Maryland	30.7%	40.7%	28.6%	2.7M
41	California	30.9%	40.6%	28.5%	18.0M
42	Alaska	34.9%	36.6%	28.5%	313K
43	Illinois	29.6%	42.0%	28.5%	6.0M
44	Rhode Island	31.9%	39.9%	28.2%	473K
45	Louisiana	33.2%	39.0%	27.9%	1.9M
46	Connecticut	31.8%	40.4%	27.8%	1.7M
47	Pennsylvania	32.8%	39.5%	27.7%	6.0M
48	District of Columbia	26.2%	46.5%	27.4%	685K
49	New York	35.6%	37.2%	27.2%	9.5M
50	Wyoming	35.7%	37.2%	27.1%	274K
51	Massachusetts	33.8%	39.9%	26.3%	3.6M

State estimates derived from BLS Occupational Employment and Wage Statistics (May 2024) cross-referenced with JobZone zone classifications. Employment counts represent jobs, not individuals. Zone percentages reflect the occupational composition of each state’s workforce.

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by United Kingdom

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

The UK's services sector (83%) is the largest of any economy we track, but its composition differs from the US — a stronger NHS and public-sector healthcare workforce provides GREEN protection. Financial services, retail, and public-sector admin still drive significant RED exposure at levels comparable to (though slightly below) the US.

All 12 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	London	31.4%	33.9%	34.7%	5.3M
2	South East	34.6%	33.7%	31.7%	4.0M
3	East of England	35.2%	34.0%	30.9%	2.8M
4	South West	35.5%	34.0%	30.5%	2.5M
5	North West	35.7%	34.1%	30.2%	3.3M
6	Scotland	35.9%	33.9%	30.2%	2.5M
7	North East	36.2%	34.1%	29.7%	1.1M
8	Yorkshire & Humber	36.1%	34.2%	29.7%	2.5M
9	West Midlands	36.4%	34.1%	29.5%	2.6M
10	East Midlands	36.5%	34.2%	29.3%	2.1M
11	Wales	36.4%	34.4%	29.2%	1.4M
12	Northern Ireland	36.7%	34.3%	29.0%	820K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Europe

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

Europe's lower services share (73%) and stronger manufacturing base shift more workers into YELLOW — roles that will be augmented rather than replaced. Countries like Germany and Italy retain significant industrial employment, which sits in the adaptation zone. Agriculture (3.3%) and traditional trades also contribute to a broader GREEN base.

All 29 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Luxembourg	34.0%	35.0%	31.0%	275K
2	United Kingdom	35.0%	35.0%	30.0%	29.0M
3	Netherlands	35.0%	35.0%	30.0%	9.2M
4	France	36.0%	35.0%	29.0%	27.0M
5	Belgium	36.0%	35.0%	29.0%	4.6M
6	Sweden	36.0%	35.0%	29.0%	5.0M
7	Switzerland	36.0%	35.0%	29.0%	4.7M
8	Denmark	36.0%	35.0%	29.0%	2.8M
9	Norway	36.0%	35.0%	29.0%	2.6M
10	Ireland	36.0%	35.0%	29.0%	2.3M
11	Cyprus	36.0%	35.0%	29.0%	360K
12	Germany	37.0%	35.0%	28.0%	45.0M
13	Spain	37.0%	35.0%	28.0%	18.5M
14	Greece	37.0%	35.0%	28.0%	3.4M
15	Finland	37.0%	35.0%	28.0%	2.4M
16	Italy	38.0%	35.0%	27.0%	20.5M
17	Portugal	38.0%	35.0%	27.0%	4.5M
18	Austria	38.0%	35.0%	27.0%	4.1M
19	Croatia	38.0%	35.0%	27.0%	1.4M
20	Lithuania	38.0%	35.0%	27.0%	1.2M
21	Latvia	38.0%	35.0%	27.0%	780K
22	Estonia	38.0%	35.0%	27.0%	620K
23	Poland	39.0%	35.0%	26.0%	15.5M
24	Hungary	39.0%	35.0%	26.0%	4.0M
25	Bulgaria	39.0%	35.0%	26.0%	2.5M
26	Slovakia	39.0%	35.0%	26.0%	2.3M
27	Slovenia	39.0%	35.0%	26.0%	900K
28	Czechia	40.0%	35.0%	25.0%	4.9M
29	Romania	41.0%	35.0%	24.0%	7.5M

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Germany

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

Germany's industrial strength — 26% of employment vs. 19% in the US — is the key differentiator. Manufacturing, engineering, and skilled trades roles sit mostly in YELLOW and GREEN. However, Germany's large admin and financial-services sector still generates significant RED exposure, particularly in banking and insurance hubs.

All 16 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Berlin	33.1%	34.1%	32.8%	1.8M
2	Hamburg	33.1%	34.4%	32.5%	1.0M
3	Hessen	34.2%	34.6%	31.2%	3.3M
4	Bayern	34.5%	34.8%	30.7%	7.4M
5	Bremen	34.7%	34.7%	30.6%	320K
6	Baden-Württemberg	34.7%	34.7%	30.5%	5.9M
7	Nordrhein-Westfalen	35.0%	34.9%	30.1%	8.9M
8	Rheinland-Pfalz	35.7%	35.3%	29.0%	2.0M
9	Niedersachsen	35.9%	35.2%	29.0%	4.0M
10	Schleswig-Holstein	35.5%	35.7%	28.8%	1.4M
11	Sachsen	36.0%	35.9%	28.1%	2.0M
12	Thüringen	36.4%	35.7%	27.9%	1.0M
13	Saarland	36.7%	35.4%	27.9%	480K
14	Brandenburg	36.2%	35.9%	27.9%	1.1M
15	Sachsen-Anhalt	36.5%	35.9%	27.6%	1.0M
16	Mecklenburg-Vorpommern	36.4%	36.1%	27.5%	750K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Japan

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

Japan has the smallest RED zone among major economies. Its unique combination of a large manufacturing sector (23%), strong robotics industry, and cultural emphasis on job retention means more workers sit in GREEN and YELLOW. Lifetime employment norms slow displacement even in at-risk roles, though this also slows adaptation.

All 47 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Tokyo	32.9%	34.9%	32.2%	7.7M
2	Osaka	34.6%	35.6%	29.8%	4.2M
3	Kanagawa	34.5%	35.8%	29.8%	4.3M
4	Saitama	34.5%	35.7%	29.8%	3.4M
5	Kyoto	34.5%	35.7%	29.7%	1.2M
6	Chiba	34.5%	35.8%	29.7%	2.9M
7	Hyogo	35.0%	35.9%	29.2%	2.5M
8	Aichi	34.9%	35.9%	29.1%	3.8M
9	Fukuoka	35.0%	35.9%	29.1%	2.4M
10	Miyagi	35.2%	35.9%	28.9%	1.1M
11	Hiroshima	35.2%	36.2%	28.6%	1.3M
12	Ishikawa	35.0%	36.7%	28.3%	540K
13	Ibaraki	35.3%	36.4%	28.3%	1.4M
14	Hokkaido	35.5%	36.2%	28.3%	2.4M
15	Nara	35.4%	36.3%	28.2%	570K
16	Tochigi	35.4%	36.5%	28.0%	920K
17	Gunma	35.4%	36.5%	28.0%	920K
18	Shizuoka	35.6%	36.3%	28.0%	1.8M
19	Okinawa	36.2%	35.8%	28.0%	650K
20	Okayama	35.6%	36.4%	28.0%	890K
21	Shiga	35.5%	36.5%	28.0%	665K
22	Fukushima	36.0%	36.4%	27.6%	870K
23	Kagawa	35.8%	36.7%	27.6%	450K
24	Toyama	35.6%	36.9%	27.5%	520K
25	Mie	36.1%	36.4%	27.5%	840K
26	Niigata	36.2%	36.4%	27.5%	1.1M
27	Shimane	35.8%	36.8%	27.4%	310K
28	Gifu	36.0%	36.6%	27.4%	950K
29	Nagano	35.9%	36.7%	27.3%	980K
30	Iwate	36.1%	36.6%	27.3%	590K
31	Fukui	36.4%	36.4%	27.3%	385K
32	Yamanashi	36.0%	36.8%	27.3%	400K
33	Kumamoto	36.1%	36.6%	27.3%	800K
34	Tokushima	36.1%	36.7%	27.2%	335K
35	Wakayama	36.2%	36.7%	27.1%	425K
36	Oita	36.3%	36.7%	27.0%	540K
37	Yamaguchi	36.3%	36.8%	26.9%	620K
38	Ehime	36.3%	36.8%	26.9%	620K
39	Saga	36.3%	36.8%	26.9%	375K
40	Tottori	36.2%	36.9%	26.9%	260K
41	Aomori	36.6%	36.6%	26.8%	590K
42	Miyazaki	36.5%	36.7%	26.7%	490K
43	Yamagata	36.3%	36.9%	26.7%	520K
44	Nagasaki	36.4%	36.9%	26.7%	585K
45	Kagoshima	36.5%	36.9%	26.6%	740K
46	Kochi	36.6%	36.9%	26.5%	325K
47	Akita	36.8%	36.8%	26.4%	440K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Canada

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

Canada closely mirrors the US pattern — a large services sector (80%) creates similar RED exposure through retail, admin, and customer-service roles. Strong natural-resources and healthcare sectors provide GREEN protection, but the composition is almost identical to the US breakdown.

All 13 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	British Columbia	34.7%	32.8%	32.5%	2.6M
2	Ontario	34.9%	32.8%	32.2%	7.7M
3	Northwest Territories	32.0%	36.0%	32.0%	25K
4	Alberta	34.9%	33.2%	31.9%	2.4M
5	Quebec	35.1%	33.0%	31.9%	4.3M
6	Yukon	36.4%	31.8%	31.8%	22K
7	Nova Scotia	35.2%	33.5%	31.3%	460K
8	Saskatchewan	35.5%	33.4%	31.1%	560K
9	Manitoba	35.8%	33.3%	30.9%	660K
10	New Brunswick	35.6%	33.6%	30.8%	360K
11	Prince Edward Island	34.6%	34.6%	30.8%	78K
12	Nunavut	38.5%	30.8%	30.8%	13K
13	Newfoundland & Labrador	35.9%	33.6%	30.5%	220K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Australia

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

Australia's economy is structurally similar to Canada and the US, with a dominant services sector (78%). Mining and natural resources provide some GREEN insulation, but the large retail, hospitality, and financial-services workforce drives RED percentages comparable to the UK.

All 8 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Australian Capital Territory	32.1%	32.1%	35.8%	240K
2	Northern Territory	35.0%	32.5%	32.5%	120K
3	Victoria	34.9%	32.8%	32.3%	3.5M
4	New South Wales	34.7%	33.0%	32.3%	4.1M
5	Western Australia	34.9%	33.2%	31.9%	1.4M
6	Queensland	35.5%	33.3%	31.2%	2.6M
7	South Australia	35.6%	33.5%	30.9%	860K
8	Tasmania	36.4%	33.5%	30.2%	242K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by South Korea

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

South Korea's significant manufacturing sector (24%) — especially electronics, automotive, and shipbuilding — places more workers in the YELLOW adaptation zone. Higher agriculture (5.2%) compared to other advanced economies adds a small GREEN buffer. The tech-forward workforce means adaptation is faster but displacement risk in service roles remains.

All 17 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Seoul	34.5%	34.1%	31.5%	5.3M
2	Gyeonggi	35.6%	34.8%	29.6%	7.0M
3	Daejeon	35.8%	34.7%	29.4%	720K
4	Sejong	36.1%	35.0%	28.9%	180K
5	Incheon	36.1%	35.1%	28.8%	1.4M
6	Busan	36.4%	35.2%	28.4%	1.6M
7	Gwangju	36.6%	35.1%	28.2%	680K
8	Daegu	36.6%	35.2%	28.2%	1.1M
9	South Chungcheong	36.5%	35.7%	27.8%	1.0M
10	Ulsan	36.9%	35.5%	27.6%	510K
11	South Gyeongsang	36.9%	35.7%	27.4%	1.5M
12	North Chungcheong	37.2%	35.5%	27.3%	740K
13	Gangwon	37.3%	35.8%	26.9%	670K
14	North Gyeongsang	37.5%	35.8%	26.7%	1.2M
15	North Jeolla	37.6%	35.9%	26.5%	800K
16	Jeju	37.4%	36.1%	26.5%	310K
17	South Jeolla	37.9%	36.2%	25.9%	810K

AI Displacement by Region

% workforce in Red Zone

Loading map…

AI Displacement by Global

How many workers in your region face AI displacement? Select a region to see its zone breakdown.

Why does this breakdown look like this?

The global average has a larger GREEN zone and significantly smaller RED zone than advanced economies. Agriculture employs 26% of the world's workers in roles with physical and environmental complexity that resists near-term automation. Lower-income economies have less services-sector employment, which substantially reduces RED exposure — consistent with IMF estimates that low-income countries face roughly half the AI displacement risk of advanced economies.

All 30 Ranked by RED Zone %

#	Region	Green	Yellow	Red	Workers
1	Canada	35.0%	33.0%	32.0%	19.0M
2	Australia	35.0%	33.0%	32.0%	13.1M
3	United Kingdom	35.0%	34.0%	31.0%	29.0M
4	France	35.0%	34.0%	31.0%	27.0M
5	Italy	35.0%	34.0%	31.0%	20.5M
6	Spain	35.0%	34.0%	31.0%	18.5M
7	Poland	35.0%	34.0%	31.0%	15.5M
8	Netherlands	35.0%	34.0%	31.0%	9.2M
9	Sweden	35.0%	34.0%	31.0%	5.0M
10	United States	29.0%	40.0%	31.0%	155.0M
11	Germany	35.0%	35.0%	30.0%	45.0M
12	Japan	35.0%	36.0%	29.0%	63.0M
13	South Korea	36.0%	35.0%	29.0%	25.5M
14	China	50.0%	28.0%	22.0%	760.0M
15	Mexico	52.0%	27.0%	21.0%	55.0M
16	Saudi Arabia	52.0%	27.0%	21.0%	11.0M
17	Malaysia	52.0%	27.0%	21.0%	14.0M
18	Brazil	52.0%	28.0%	20.0%	90.0M
19	Russia	50.0%	30.0%	20.0%	66.0M
20	Turkey	52.0%	28.0%	20.0%	30.0M
21	Thailand	52.0%	28.0%	20.0%	35.0M
22	Philippines	58.0%	24.0%	18.0%	42.0M
23	South Africa	58.0%	24.0%	18.0%	14.0M
24	Egypt	60.0%	24.0%	16.0%	27.0M
25	India	62.0%	23.0%	15.0%	500.0M
26	Indonesia	62.0%	23.0%	15.0%	125.0M
27	Vietnam	62.0%	23.0%	15.0%	47.0M
28	Nigeria	64.0%	22.0%	14.0%	50.0M
29	Pakistan	64.0%	22.0%	14.0%	50.0M
30	Bangladesh	66.0%	21.0%	13.0%	67.0M

What the Data Shows

JobZone Risk scores 3,649 professions across 28 industry domains, covering 100% of the US workforce. Each role is rated 0–100 on an 8-dimension scoring framework. The average score is 45.1 out of 100.

Role coverage has been cross-validated against five international occupational classification systems (O*NET, ESCO, UK SOC 2020, Canadian NOC, ANZSCO) — 6,427 classified occupations, 100% match rate.

Roles scoring below 25 are classified RED (high displacement risk), 25–47 are YELLOW (transforming), and 48+ are GREEN (AI-resistant).

AI Resistance by Industry Domain

Trades, healthcare, and emergency services score highest. Business operations, IT support, and data entry score lowest.

Trades & Physical

60.5 Veterinary & Animal Care

59.8 Military

57.6 Healthcare

57.5 Sports & Recreation

56.2 AI

56.0 Social Services

55.8 Religious & Community

54.4 Public Safety

53.0 Utilities & Energy

46.0 Government & Public Admin

42.4 Retail & Service

40.8 Science & Research

40.7 Legal & Compliance

39.7 Library, Museum & Archives

39.4 Creative & Media

37.2 Development

36.0 Cloud & Infrastructure

35.1 Real Estate & Property

34.5 Manufacturing

31.1 Business & Operations

29.6 Data

28.6

Who Is Most Affected

Entry-level white-collar roles face the highest displacement risk. Roles requiring physical presence, complex judgement, or deep human interaction score safest.

Top 10 Safest Roles

#	Role	Score	Zone
1	Electrical Power-Line Installer and Repairer (Mid-Level)	91.6	GREEN
2	Signalling Tester In Charge / STIC (Mid-Level)	87.7	GREEN
3	Model Alignment Researcher (Mid-Level)	86.1	GREEN
4	AI Safety Researcher (Mid-Senior)	85.2	GREEN
5	Foster Carer (Mid-Level)	84.5	GREEN
6	Leadworker (Mid-Level)	83.7	GREEN
7	Heat Pump Installer (Mid-Level)	83.5	GREEN
8	Trauma Surgeon (Mid-to-Senior)	83.2	GREEN
9	CCS Engineer (Control Command & Signalling) (Mid-Level)	83.2	GREEN
10	Chief Information Security Officer (CISO) (Senior/Executive)	83.0	GREEN

Top 10 Most At-Risk Roles

#	Role	Score	Zone
1	File Clerks (Mid-Level)	1.5	RED
2	Micro-Task Worker (Online) (Mid-Level)	1.7	RED
3	Data Entry Keyer (Mid-Level)	2.3	RED
4	Word Processor and Typist (Mid-Level)	2.6	RED
5	Vulnerability Tester / Scanner Operator (Entry-Level)	2.7	RED
6	Telephone Operator (Mid-Level)	3.0	RED
7	Virtual Assistant (Entry-to-Mid Level)	3.2	RED
8	Live Chat Support Agent (Entry-to-Mid Level)	3.4	RED
9	Telemarketer (Mid-Level)	3.4	RED
10	Medical Transcriptionist (Mid-Level)	3.6	RED

Browse all 3,649 scored roles →

It’s Already Happening

These aren’t forecasts. Real companies are already cutting real jobs and citing AI as the reason.

2026 tech layoffs reach 45000 in March, more than 9200 due to AI and automation

As of March 2026, over 9,200 of the 45,363 tech layoffs worldwide have been linked to AI implementation and organizat...

Google · 21d ago

Block spent 18 months 're-skilling staff' with AI. Now it is laying off 4,000 workers.

Google News · 27d ago

Verizon to Cut 13,000 Jobs, Set Up $20 Million “Reskilling” Fund for Laid Off Staff In “Age of AI”

Google News · 130d ago

Tech Layoffs Surge While AI Jobs Soar: Key Trends Shaping the 2026 Tech Industry

Early 2026 has seen over 45,000 global tech layoffs, largely driven by AI restructuring, even as hiring for AI-relate...

Google · 9d ago

AI is eating entry-level tech jobs, and young developers are paying the price

Employment for software developers under 25 declined nearly 20% by July 2025, while entry-level tech hiring dropped 2...

Google · 14d ago

Labor market impacts of AI: A new measure and early evidence - Anthropic

Anthropic introduces a new measure of AI displacement risk, 'observed exposure,' combining theoretical LLM capabiliti...

Google · 25d ago

Nordea plans job cuts amid restructuring initiatives - FinTech Futures

Nordea's plans are set to impact around 1,500 employees by 2027, with the restructuring expected to deliver cost savi...

Fintechfutures · 12d ago

Jamie Dimon just made a bold prediction about AI and your job - TheStreet

Jamie Dimon, CEO of JPMorgan Chase, predicts AI will lead to shorter workweeks (3.5 days) and longer, healthier lives...

Google · 25d ago

See all AI displacement news →

AI Also Creates Jobs — But Different Ones

Displacement is only half the story. AI adoption is simultaneously creating entirely new occupations that didn’t exist five years ago. The challenge for policymakers: the people losing jobs aren’t the same people getting the new ones.

New Roles Hiring Now

JobZone has assessed 30 AI-created roles. These five have real job postings, real salaries, and measurable year-over-year growth based on our REL framework analysis.

AI Safety Researcher +78% YoY AI Agent Architect +71% YoY AI Red Team Specialist +63% YoY AI Governance Lead +50% YoY AI Auditor +50% YoY

Salaries range $95K–$300K+. Regulation (EU AI Act, NIST AI RMF) is a key driver of AI governance and safety hiring. Browse all assessed roles →

What the Forecasters Say

World Economic Forum (2025)

Projects 170 million new jobs created globally by 2030, against 92 million displaced — a net gain of 78 million. (Note: WEF’s 2023 report projected a net loss of 14 million. The revision reflects faster-than-expected AI-adjacent job creation.)

McKinsey Global Institute

Long-term net impact “likely positive” due to productivity gains, but up to 12 million U.S. workers may need to switch occupations by 2030. Massive reskilling required.

OECD Employment Outlook

Negative employment effects from AI have not yet materialised to a significant extent. 27% of OECD jobs at high risk. Outcome depends heavily on policy actions.

Anthropic Research (March 2026)

First-party AI usage data shows no systematic unemployment increase for highly exposed workers yet — but hiring of workers aged 22–25 has dropped ~14% in exposed occupations since 2022. AI currently covers only 33% of theoretically automatable tasks in Computer & Math roles, suggesting a narrowing but real policy window before displacement accelerates. Their 10 most-exposed occupations directly correlate with JobZone Risk zone classifications — 10 out of 10 confirmed. Source →

Indeed Hiring Lab (Jan 2026)

AI job postings growing 130%+ year-over-year, but in an otherwise weak “low-hire, low-fire” labour market. AI roles are bucking the trend.

Why this doesn’t solve the problem

Net job creation projections assume successful reskilling at scale. The new roles require significant technical skills, pay $100K+, and concentrate in metros with existing tech ecosystems. A displaced call centre worker in Ohio cannot become an AI Safety Researcher without years of retraining and likely relocation. The policy implication: job creation numbers are meaningless without transition infrastructure.

Sources: WEF Future of Jobs 2025, McKinsey, OECD, Indeed Hiring Lab, PwC AI Jobs Barometer. AI-created role data from JobZone’s own REL framework assessment of 30 roles.

How Fast Is This Moving?

The question isn’t whether AI will reshape the labour market — it’s how quickly different sectors will be hit. Institutional forecasts and our own JobZone data converge on three overlapping phases.

Phase 1: Now

2024–2027

Already underway

Knowledge-work automation via LLMs. Entry-level white-collar displacement accelerating. First factory humanoid deployments.

Who’s affected:

RED-zone roles — data entry, bookkeeping, junior developers, admin assistants, SDRs.

Phase 2: Near-Term

2027–2032

Accelerating

Agentic AI handling multi-step workflows. Semi-structured robotics pilots in hospitals and new construction. Professional services disruption.

Who’s affected:

YELLOW-zone roles — analysts, mid-level tech, paralegals, some professional services.

Phase 3: Medium-Term

2032–2040+

Physical work enters

Robotics expands into unstructured environments — homes, existing buildings, bedside care. Task automation begins in trades and healthcare.

Who’s affected:

Currently GREEN roles with physical-world protection start eroding — but skilled trades and bedside care retain 20–35+ years of protection.

What the Data Shows

Entry-level employment in AI-exposed roles down ~13% since 2022 (up to 16% relative to less-exposed occupations). Entry-level software development job postings down nearly 20% from their late-2022 peak. The pipeline damage may be irreversible within 2–3 years without intervention.

Stanford Digital Economy Lab / Brynjolfsson et al.; SPARK6

300 million full-time jobs globally exposed to generative AI. Two-thirds of US occupations have some exposure, with 25–50% of workload potentially replaceable.

Goldman Sachs (March 2023)

40% of global employment exposed to AI — rising to 60% in advanced economies. In low-income countries, 26%. Of exposed jobs in advanced economies, roughly half may benefit; the other half face reduced demand.

IMF (January 2024)

50% of work activities could be automated between 2030 and 2060, with a midpoint of ~2045. Generative AI accelerated this estimate by roughly a decade compared to McKinsey’s pre-2023 forecast.

McKinsey Global Institute (June 2023)

PwC’s Three Waves: Algorithm wave (~3% of jobs, to early 2020s) → Augmentation wave (~20%, to late 2020s) → Autonomy wave (~30%, to mid-2030s). Each wave expands the blast radius from data tasks to physical-world tasks.

PwC (February 2018)

Both sides are accelerating

The gap between forecasts and reality is shrinking. WEF revised its 2023 projection (net −14 million jobs by 2027) to net +78 million by 2030 — not because displacement slowed, but because new-role creation is happening faster than expected. Displacement and creation are both accelerating. The policy implication: the window for proactive intervention is narrower than it appears.

Sources: PwC Three Waves (2018), Goldman Sachs (2023), IMF (2024), McKinsey (2023), WEF Future of Jobs 2025. AIJRI zone data and robotics convergence timelines from JobZone’s own methodology.

What Governments Are Already Doing

United States

AI-Related Job Impacts Clarity Act (S.3108) — Bipartisan bill by Senators Warner and Hawley requiring companies to submit quarterly reports on AI-driven workforce changes to the Department of Labor. First mandatory national data collection on AI’s employment effects.
No Robot Bosses Act (H.R.6371) — Requires employers with 11+ employees to audit AI tools for bias, provide human oversight of AI decisions, and offer human appeal of adverse AI decisions.
DOL AI Literacy Framework (Feb 2026) — Five content areas for workforce AI literacy: understanding AI principles, exploring uses, directing AI, evaluating outputs, responsible use.
State action — 38 states adopted or enacted AI-related legislation in 2025. Colorado’s AI Act (SB24-205, effective June 2026) requires impact assessments and human review of adverse decisions. Illinois requires employer notification when AI is used in employment decisions.

United Kingdom

AI Regulation Bill — Private member’s bill re-introduced to the House of Lords by Lord Holmes (March 2025), proposing a dedicated AI Authority as a regulatory body.
National AI Competency Expansion — Government-industry programme to provide 10 million workers with AI skills by 2030, building on the AI Opportunities Action Plan.
UBI discussion — Lord Stockwood (Minister for Investment) has publicly discussed the need for “some sort of UBI, some sort of lifelong learning mechanism” to support workers in AI-threatened industries, and has proposed funding through tech company taxation.

European Union

EU AI Act — Employment-related AI classified as “high-risk” from August 2026, requiring conformity assessments, risk management, human oversight, transparency, and data governance before deployment.

Other Notable Approaches

Singapore — National AI Strategy 2.0 with substantial resources for talent and ethics. Ranked #1 on the Coursera AI Maturity Index (2025).
South Korea — 1.4 trillion won (~$1 billion) investment in AI talent development for 2026. Training 11,000 high-level AI specialists. AI Education Support Centres expanding to all 17 regional education offices by 2028.
Denmark — Flexicurity model combining flexible labour markets with strong social safety nets, whose inherent flexibility naturally supports AI-era workforce transitions.
Nordic countries — NordForsk cross-border AI research fund with ethical guidelines. Finland dramatically reformed lifelong learning systems to integrate AI skills.

Evidence-Based Policy Options

10 policy levers grouped by implementation timeline. Each is drawn from real-world proposals, legislation, or expert recommendations.

Short-Term (0–12 months)

1 Mandatory AI Displacement Reporting

Require companies to report AI-driven workforce changes quarterly. You cannot manage what you cannot measure.

Real-world example: The bipartisan Warner-Hawley AI-Related Job Impacts Clarity Act (S.3108) would create the first mandatory national dataset on AI’s employment effects.

Data tie-in: JobZone Risk already scores 3,649 roles. Government reporting would validate and extend this data with actual displacement figures.

Estimated cost: ~$54,000/year per firm. No AI-specific reporting cost data exists yet, so this estimate is based on analogous SEC quarterly reporting compliance costs for registered advisers. The administrative burden of quarterly data collection and submission to a federal agency is structurally similar. Actual costs will vary significantly by company size.

2 Advance Notice Requirements (WARN Act Update)

Extend WARN Act mass-layoff notification requirements to cover AI-driven workforce reductions, even when they occur gradually rather than all at once.

Real-world example: Illinois requires employer notification when AI is used in employment decisions. Colorado’s AI Act (effective June 2026) will add similar requirements.

Data tie-in: Our data shows 44.3M workers in RED-zone roles. Advance notice gives workers time to retrain before displacement hits.

Estimated cost: Compliance cost is primarily administrative (drafting and distributing notices). The real cost is non-compliance: up to ~$9,600 per affected employee in back pay plus $500/day in civil penalties. For a 50-person layoff, a 60-day violation could cost ~$480,000 in back pay alone. Extending WARN to AI-driven reductions adds no new cost category — just a broader trigger.

3 AI Displacement Insurance

Parametric insurance that pays out automatically when a worker’s role crosses a displacement threshold. Up to 50% of net pay on every scheduled payday for 3–12 months.

Real-world example: Singularity AI Job Loss Insurance already offers this as a commercial product with dual-trigger mechanisms.

Data tie-in: JobZone scores could serve as one trigger in the parametric model, creating an objective, data-driven activation mechanism.

Estimated cost: Unknown. Singularity’s product exists commercially but premiums are not publicly disclosed. The AI displacement insurance market is too new for benchmark pricing — no actuarial data on AI-driven job loss frequency exists yet, making premium modelling speculative.

Medium-Term (1–3 years)

4 AI Employment Impact Assessments

Require pre-deployment impact assessments for any AI system affecting hiring, firing, promotion, or scheduling decisions.

Real-world example: Colorado’s AI Act (SB24-205, effective June 2026) requires annual impact assessments for high-risk employment AI. The EU AI Act mandates conformity assessments from August 2026.

Data tie-in: Our domain-level data shows which sectors face the greatest transformation. Assessments can be prioritised based on sector risk profiles.

Estimated cost: $2,400–$500,000/year per company for compliance tools and services under Colorado’s AI Act. The wide range reflects company size: small firms can use SaaS platforms from ~$199/month, while enterprise compliance requires $50K–$500K/year. Non-compliance penalties start at $20,000 per violation, compounding to $100K–$5M+ per incident.

5 Lifelong Learning Accounts

Tax-deferred retraining savings accounts, similar to retirement accounts, that workers can draw on for approved reskilling programmes.

Real-world example: Brookings Institution recommends lifelong learning accounts with tax-deferred savings for worker retraining. Singapore’s SkillsFuture programme provides every citizen with credits for lifelong learning.

Data tie-in: Workers in YELLOW-zone roles (68.1M workers) are the prime target — their roles are transforming and reskilling now prevents displacement later.

Estimated cost: $500 per citizen (base credit) plus $4,000 mid-career top-up for workers aged 40+, based on Singapore’s SkillsFuture — the only fully operational national learning account system. Enterprise credits reach up to $10,000 per worker. Course fee subsidies cover 50–95% depending on age and need. Singapore is the best cost benchmark because no other country has implemented this at national scale.

6 Sector Transition Funds

Targeted government funds for sectors facing the highest AI disruption, providing wage subsidies, retraining grants, and entrepreneurial support during transition.

Real-world example: Gene Sperling’s Economic Dignity Compact proposes $10,000 wage bonuses for displaced workers and training support for 50 million people annually. South Korea is investing ~$1 billion in AI talent development for 2026.

Data tie-in: Our domain averages reveal which sectors need intervention most urgently — business operations and IT support average deep into the YELLOW and RED zones.

Estimated cost: ~€20,000–€50,000 per displaced worker, based on the EU Just Transition Fund (€17.5 billion total, 2021–2027) — the closest real-world precedent for sector-specific displacement support at scale. That fund targeted coal regions; AI displacement is broader and more diffuse, so per-worker costs could differ substantially. South Korea’s ~$1.2 billion (2026) focuses on building an AI talent pipeline rather than direct worker transition.

7 Portable Benefits

Detach health insurance, retirement, and other benefits from specific employers so workers changing careers due to AI are not penalised twice — losing both their job and their safety net.

Real-world example: Brookings recommends passing the Portable Benefits for Independent Workers Pilot Program Act and reducing retirement vesting requirements from 3–6 years.

Data tie-in: Workers transitioning from RED-zone to GREEN-zone roles (e.g., data entry clerk → cybersecurity analyst) should not lose healthcare coverage during the switch.

Estimated cost: The Pilot Act authorises $20 million for testing. Existing models range widely: gig platforms contribute ~$31/month per worker (~4% of earnings), while union-negotiated portable benefits run 11–11.5% of gross compensation. The gap reflects the difference between minimum gig-economy contributions and comprehensive coverage.

Long-Term (3–10 years)

8 Automation Tax / Robot Services Tax

Tax automated services (AI customer support, robotic delivery) rather than robots themselves — preserving investment incentives while generating revenue to fund displaced worker support.

Real-world example: Brookings advocates a “robot services tax” model. Sen. Sanders proposed a levy for each human position replaced. Nobel laureate Daron Acemoglu recommends equalising capital and labour tax rates to remove the artificial subsidy for automation.

Data tie-in: The data can quantify the scale of potential tax revenue by showing how many workers in each sector face displacement.

Estimated cost: MIT economists recommend an optimal tax rate of 1–3.7% of robot/AI system value. However, no government body — not the CBO, Brookings, or anyone else — has published projected revenue figures for a US automation tax. This is genuinely uncharted policy territory, making cost-benefit analysis impossible with current data.

9 UBI Pilots / Income Support

Pilot universal basic income or targeted income support for workers in high-displacement sectors. Multiple proposals exist with different funding mechanisms.

Real-world example: UK Investment Minister Lord Stockwood has publicly discussed UBI funded by tech company taxation. Andrew Yang advocates a $1,000/month Freedom Dividend. Sam Altman’s American Equity Fund would tax companies 2.5% of market value annually. Gene Sperling’s Economic Dignity Compact proposes 20 million subsidised high-purpose jobs at $75,000 each as an alternative to UBI.

Data tie-in: 724.5M workers globally are in RED-zone roles. The scale of potential displacement can inform the size and targeting of income support pilots.

Estimated cost: The range is enormous. Finland’s 2,000-person pilot cost ~€25 million over 2 years. Yang’s universal $1,000/month Freedom Dividend would cost ~$2.8 trillion/year (~12–14% of GDP). Sperling’s targeted alternative claims savings of “~$1 trillion vs UBI.” The massive range reflects fundamental design choices: universal vs targeted, payment level, and duration. Small pilots are cheap; national programmes are among the most expensive policies ever proposed.

10 Education System AI Literacy Mandates

Embed AI literacy across all K–12 subjects (not just computer science) and reform higher education to prioritise skills that complement AI rather than compete with it.

Real-world example: 28 US states have published AI guidance for K–12 education. Vermont’s framework has age-appropriate tiers. South Korea is establishing AI Education Support Centres at all 17 regional education offices by 2028.

Data tie-in: Our data shows physical, interpersonal, and judgement-heavy skills score GREEN. Education should steer toward roles where humans have a durable advantage — use our Compare tool to see specific transition paths.

Estimated cost: South Korea committed 900 billion won (~$750 million) to K–12 AI education in 2026 — the only country to have published a dedicated AI education budget. No per-student cost data exists anywhere globally. Implementation costs depend heavily on whether AI literacy is woven into existing subjects (cheaper) or requires new dedicated instruction and teacher retraining (significantly more expensive).

Access the Data

All JobZone Risk data is free. Search roles, explore dashboards, or plug the full dataset directly into your own analysis.

Search any role

Look up any of 3,649+ roles and see AI displacement scores, zone ratings, and task-level breakdowns.

Live data monitor

Real-time dashboard with country breakdowns, domain analysis, and zone distribution across the full workforce.

Compare & transition paths

Compare at-risk roles against safer alternatives to identify practical retraining pathways for displaced workers.

Full dataset for LLMs

The complete database in plain text — optimised for ChatGPT, Claude, and other AI assistants. Feed it in and ask questions about any role.

Embeddable widgets

Drop a live risk-score widget into your own site or report. Copy one line of HTML — scores update automatically as new data comes in.

Need something else? For bulk data exports, custom analysis, or integration support — get in touch.

About the Data

Every role receives a JobZone Score (0–100) based on an 8-dimension framework measuring current AI capability overlap, automation barriers, displacement evidence, protective factors, market dynamics, and growth correlation. Based on 47 data sources including Oxford, McKinsey, OECD, BLS, and Anthropic research. Full methodology →

Scores are not static. JobZone Risk continuously monitors global news feeds, tracking AI capability announcements, labour market shifts, and regulatory changes. When a material event occurs, affected role scores are re-evaluated and updated.

Nathan House is a cybersecurity and AI expert with 30 years of hands-on experience. He has secured systems at companies including Vodafone, BP, ExxonMobil, and GSK, and trained organisations including Microsoft, Cisco, Siemens, and Thales. He built JobZone Risk after realising AI would displace the very entry-level cybersecurity jobs his 500,000+ students were training for — and that no rigorous, data-driven index existed to quantify the risk. He is founder of StationX, one of the world’s largest cybersecurity training platforms. Why I built this →

Awards & Recognition

Media

CNN, Fox News, NBC, NDTV
PC Extreme Magazine
Published white papers on AI, cybersecurity trends, and emerging threats

Certifications

CISSP, OSCP, CEH, CISM, CISA, ISO 27001 Lead Auditor, SABSA

Connect

55K+ LinkedIn followers

Suggested Citation

House, N. (2026). AI Job Resistance Index (AIJRI v3.2). JobZone Risk. https://jobzonerisk.com/policymakers

Contact

Nathan House is available for policy consultations, expert testimony, and data requests.

[email protected] WhatsApp @GotoNathan on X

Typical response time: same business day.