AI May Replace 1.7 Million Bangladesh RMG Workers, But Create 800,000 New Tech Jobs by 2035

Dhaka, August 11, 2026 — A new study on artificial intelligence and the future of work in Bangladesh’s ready-made garment (RMG) sector has concluded that up to 1.7 million of the country’s 4.2 million garment workers could face job displacement by 2035 as AI-driven cutting, sewing, inspection and packaging systems move from pilot to industrial scale — but the same transition is projected to create 800,000 new jobs in quality assurance, data analytics, robotics maintenance, product design and supply-chain software.

🤖 The Study and Its Methodology

The study, released on August 10 by the Centre for Policy Dialogue (CPD) in collaboration with the International Labour Organization (ILO) and BGMEA, modelled three adoption scenarios — slow, moderate and rapid — across nine RMG sub-processes: fabric spreading, cutting, sewing, embroidery, washing, finishing, quality inspection, packing, and warehouse logistics. Under the moderate adoption scenario (the study’s central case), AI and robotics penetration reaches 38 per cent of RMG process steps by 2030 and 64 per cent by 2035.

The study draws on technology adoption data from Vietnam, China, Turkey and Sri Lanka, BMET skills registry data, factory-level surveys of 312 BGMEA member units, and procurement records from leading automation vendors including Lectra, Gerber, Juki, Brother and Pegasus.

📊 Where Job Losses Will Hit Hardest

The study identifies the five RMG sub-processes most exposed to automation:

• ✂️ Cutting and spreading: 92 per cent of surveyed large factories (500+ workers) have already deployed computer-controlled spreading and CNC cutting. Smaller factories are catching up, with 67 per cent planning adoption by 2028. Estimated direct job loss: 320,000 cutting-room workers by 2035
• 👁️ Quality inspection: AI vision systems now detect fabric defects, seam misalignment and colour mismatch at 99.2 per cent accuracy — well above the 89–93 per cent human benchmark. Estimated displacement: 240,000 inline and end-of-line inspectors
• 🧹 Washing and finishing: Automated washing lines with AI-controlled chemical dosing reduce labour by 55 per cent per unit. Estimated displacement: 180,000 workers in washing plants
• 📦 Packing and warehouse logistics: Robotic pick-and-place systems, conveyor sortation and AGV (automated guided vehicle) movement now handle 41 per cent of warehouse tasks in modern factories. Estimated displacement: 280,000 workers
• 🧵 Embroidery and basic sewing: Computerised multi-head embroidery and programmable sewing units are replacing semi-skilled operators in basic product categories (T-shirts, polo shirts, basic trousers). Estimated displacement: 680,000 operators

Critically, the study finds that displacement will be heavily skewed toward women workers, who hold 60.8 per cent of the operator-level jobs in the most automatable sub-processes. Without targeted reskilling, the gendered impact of AI in Bangladesh’s RMG sector could reverse two decades of female labour force participation gains.

💡 The Other Side: New Jobs Created

The study’s more optimistic finding is that AI adoption will create entirely new categories of skilled employment within and adjacent to the RMG sector. By 2035, the study projects the following new-job creation in Bangladesh:

• 💻 Quality assurance technologists: 180,000 new roles to operate, calibrate and audit AI vision systems and to handle complex cases the AI escalates
• 🔧 Robotics and automation maintenance engineers: 140,000 roles servicing cutting robots, AGVs, automated sewing cells and smart warehouses
• 📈 Data analysts and operations researchers: 110,000 roles in factories and HQs to interpret AI production data, optimise line balancing, and forecast demand
• 🎨 Product designers and digital pattern engineers: 95,000 roles using 3D design software, virtual sampling and AI-assisted trend analysis
• 📱 IT, cybersecurity and software roles: 180,000 roles across factories, buyers, ERP providers and supply-chain platforms
• 📊 Sustainability and traceability specialists: 95,000 roles to manage carbon accounting, materials traceability and ESG reporting now required by EU and US buyers

Net impact: a loss of 1.7 million operator-level jobs offset by 800,000 new skilled jobs — a net displacement of approximately 900,000 workers, concentrated among the least-skilled and most-female segment of the workforce.

💬 BGMEA Response: Reskilling Fund

BGMEA President Mahmud Hasan Khan (Babu), whose organisation co-sponsored the study, called for an industry-wide reskilling fund financed jointly by factory contributions, buyer commitments and government matching grants. “AI is not coming — it is already here,” he said at the study’s launch. “The question is whether we use the next nine years to retrain our workers for the new jobs, or whether we wait until 2032 and try to manage the displacement crisis. The first option is humane, smart and competitive. The second is a tragedy we will regret for a generation.”

The proposed reskilling fund would be structured as follows:

• 💰 Factory contribution: 0.15 per cent of export earnings (approximately $70 million per year at current export levels)
• 💰 Buyer contribution: $0.05 per garment from the top 50 international brands sourcing from Bangladesh, estimated at $50–$60 million per year
• 💰 Government matching: Tk 800 crore per year from the Skills Development Fund, contingent on factory and buyer contributions materialising
• 💰 Development partner grants: ILO, World Bank and ADB have signalled interest in co-financing the first three years of the programme

The fund would target retraining for 200,000 workers per year, focusing on women aged 25–45 who are most at risk of displacement and least likely to find alternative formal employment without intervention.

🌏 The Global Competitive Context

The CPD–ILO–BGMEA study also benchmarked Bangladesh against five competitor RMG exporters. The findings are sobering:

• 🇻🇳 Vietnam: Already at 51 per cent AI/automation penetration; government-backed $1.2 billion reskilling programme running since 2024
• 🇨🇳 China: 78 per cent penetration; 1.4 million workers retrained through the Made in China 2025 programme
• 🇹🇷 Turkey: 46 per cent penetration; retraining via vocational high school partnerships
• 🇱🇰 Sri Lanka: 31 per cent penetration; reskilling via ILO programme
• 🇧🇩 Bangladesh: 19 per cent penetration (rising rapidly); no national reskilling programme yet in place

The competitive implication is stark: at current adoption rates, Bangladesh’s cost-per-garment advantage of $0.18–$0.32 over Vietnam will erode to near-zero by 2030 unless automation and skills investment accelerate together.

✅ What Comes Next

BGMEA has convened a 15-member AI and Future of Work committee, co-chaired by the Ministry of Labour and Employment and the ICT Division, to finalise the reskilling fund design by October 2026 and begin pilot training programmes in January 2027. Initial pilots will focus on the 200 factories with the highest automation adoption rates — located in Gazipur, Narayanganj and Chattogram — and will train workers in robotics maintenance, QA technology, data analytics and digital pattern design.

For Bangladesh’s 4.2 million RMG workers, the message from the study is mixed. The jobs they hold today will not all exist in 2035. But the jobs that will exist — better paid, safer, less repetitive — are within reach if the industry, the government and the buyers invest now in the transition. The cost of inaction is far higher: a generation of unskilled women workers pushed out of formal employment, an export sector losing ground to Vietnam and India, and a macroeconomic shock that could shave 0.6–0.9 percentage points off annual GDP growth for a decade.

As the CPD’s lead researcher put it: “AI is not the enemy. Inaction in the face of AI is the enemy. Bangladesh has a narrow window — perhaps five years — to get this right. After that, the curve runs away from us.”