Key findings
- Clerical workers score 8.5/10 on AI exposure - approximately 2.96 million bank clerks, data entry operators and government staff face high displacement risk
- 74 million total workers - Bangladesh has one of the youngest and fastest-growing workforces in Asia, with 2+ million new entrants per year
- Garment plant operators score 7.5/10 on robotics risk - the 4 million RMG workers who power Bangladesh's $47B export engine face structural automation pressure
- Agriculture scores 2.0/10 - 30% of the workforce is in farming, the safest group from AI, but income volatility from climate events is a separate concern
The most AI-exposed occupations in Bangladesh
Bangladesh's AI risk is concentrated in the urban formal sector - particularly in Dhaka's banking and finance district, government administration, and the small but growing tech and outsourcing industry.
| Occupation group (ISCO-08) | AI score | Robotics score | Workers | Share of workforce |
|---|---|---|---|---|
| Clerical support workers (4) | 8.5/10 | 2.0/10 | 2.96M | 4.0% |
| Professionals (2) | 6.5/10 | 2.0/10 | 3.70M | 5.0% |
| Technicians and associate professionals (3) | 5.5/10 | 3.0/10 | 2.96M | 4.0% |
| Service and sales workers (5) | 4.0/10 | 2.5/10 | 8.88M | 12.0% |
| Managers (1) | 4.0/10 | 1.5/10 | 1.48M | 2.0% |
| Craft and related trades workers (7) | 3.0/10 | 4.5/10 | 8.88M | 12.0% |
| Plant and machine operators (8) | 3.5/10 | 7.5/10 | 7.40M | 10.0% |
| Elementary occupations (9) | 3.0/10 | 3.5/10 | 15.54M | 21.0% |
| Skilled agricultural workers (6) | 2.0/10 | 4.0/10 | 22.20M | 30.0% |
Source: ILO ILOSTAT, Bangladesh Bureau of Statistics (BBS) Labour Force Survey 2022. AI scores reflect exposure to large language model and automation capabilities using ISCO-08 major group task analysis. Robotics scores reflect physical automation potential. RMG workers are primarily classified under Plant and machine operators (group 8).
Why the garment sector is the defining risk - not clerical workers
Bangladesh's clerical workers score the highest on AI (8.5/10), but they represent just 4% of the workforce. The story that defines Bangladesh's automation risk is in group 8 - Plant and machine operators - which scores 7.5 out of 10 on robotics risk and includes the vast majority of the country's 4 million ready-made garment workers.
Bangladesh became the world's second largest garment exporter by building a cost and scale advantage: low wages, enormous factories, and a reliable production ecosystem around Dhaka. That model is under pressure from two directions. First, wage inflation: Bangladesh's minimum wage for garment workers has roughly doubled since 2018 (reaching around 12,500 taka/month after 2023 protests), narrowing the cost gap with automated alternatives. Second, automation cost reduction: industrial sewing robots that cost $100,000+ per unit in 2015 are approaching $20,000-30,000 in some configurations, changing the payback calculation for large factory owners. Companies like Zara, H&M, and Primark - all major Bangladesh buyers - are publicly committed to supply chain automation roadmaps.
The clerical workforce faces a separate but real AI threat. Dhaka's banking district (Motijheel) and government ministries collectively employ several hundred thousand clerks. Bangladesh Bank and Dutch-Bangla Bank have both announced digital banking initiatives that reduce branch transaction volumes. The government's Digital Bangladesh and Smart Bangladesh programs are digitalising document workflows in ways that reduce clerical headcount per unit of administrative output.
The safest jobs from AI in Bangladesh
| Occupation group (ISCO-08) | AI score | Workers | Why protected |
|---|---|---|---|
| Skilled agricultural workers (6) | 2.0/10 | 22.2M | Physical outdoor tasks, smallholder farming, seasonal adaptation |
| Elementary occupations (9) | 3.0/10 | 15.5M | Physical labour, variable urban environments, very low AI ROI |
| Craft and related trades (7) | 3.0/10 | 8.9M | Manual skill, bespoke production, small-batch work |
Agricultural work scores 2.0 out of 10 on AI exposure. Bangladesh's 22.2 million farm workers - growing rice, jute, and vegetables on millions of smallholder plots across the delta - are performing tasks that require physical presence, seasonal judgment, and adaptation to the world's most flood-prone agricultural landscape. AI cannot replicate the embedded local knowledge of a Bangladeshi farmer managing cyclone-season planting decisions.
Elementary workers (15.5 million) include construction labourers, rickshaw drivers, domestic workers and street traders. These groups face economic pressure from wage competition and informality, not from algorithmic displacement. The 3.0 out of 10 score reflects that AI simply has no efficient interface to their physical work environments.
What this means for Bangladesh's workforce
Bangladesh faces an acute version of a challenge common across developing Asia: the export sector that drove 30 years of poverty reduction is becoming automatable just as the country still needs it for employment. The RMG sector lifted millions of women into formal employment - female labour force participation jumped from under 20% in 1990 to over 36% by 2022, largely on the back of garment factory jobs. If those jobs automate over the next decade, the substitution needs to come from somewhere else.
The Bangladesh government's Smart Bangladesh 2041 vision explicitly targets IT exports and digital services as the successor growth sector. The Bangladesh Hi-Tech Park Authority is developing IT parks in Dhaka, Sylhet and Jessore. The challenge is pace: Bangladesh currently exports around $800 million in IT services annually. Replacing a meaningful share of $47 billion in garment exports would require a 50-fold expansion in under two decades - an extremely ambitious trajectory even with strong policy support.
For workers in 2026, the practical risk timeline differs sharply by sector. Clerical workers in formal banking face AI-driven workload reduction within 2-3 years as document automation tools reach commercial deployment. Garment workers face a longer but structurally more serious threat: full robotic sewing is not yet at scale for the variety and volume Bangladesh handles, but by 2030-2032 the economics will have shifted significantly. Agricultural workers face the least AI pressure of any group, but the most climate pressure - cyclone frequency, salinity intrusion and flood patterns are the primary threat to their livelihoods, on a timeline independent of AI.
Explore Bangladesh's full occupation data
See detailed AI and robotics exposure scores for all occupation groups, compare with 205 other countries, and filter by sector.
Open Bangladesh in Explore toolWas this analysis useful?
Methodology
Employment data sourced from ILO ILOSTAT, based on the Bangladesh Bureau of Statistics (BBS) Labour Force Survey 2022 (released 2023). Coverage: 74 million employed workers classified by ISCO-08 major occupation groups (1-digit). AI exposure scores (0-10) reflect the proportion of tasks within each occupation group susceptible to large language model capabilities including text generation, classification, data analysis and decision support. Robotics scores reflect susceptibility to physical automation. Weighted average (3.28/10) is calculated as employment-weighted mean across all occupation groups. The RMG sector's garment workers are primarily captured in ISCO-08 group 8 (Plant and machine operators). Individual occupation scores are analytical estimates based on ILO task taxonomy research and should not be read as precise probabilities of job loss.
Frequently asked questions
Related analyses
Sources
- ILO ILOSTAT - Bangladesh employment by occupation (ISCO-08), sourced from BBS Labour Force Survey 2022 (released 2023). ilostat.ilo.org
- Bangladesh Bureau of Statistics (BBS) Labour Force Survey 2022. bbs.gov.bd
- Bangladesh Garment Manufacturers and Exporters Association (BGMEA) - RMG sector export statistics 2023. bgmea.com.bd
- Bangladesh Hi-Tech Park Authority - Smart Bangladesh 2041 digital economy targets. bhtpa.gov.bd