The government wants to spend Tk 683 crore training more than 50,000 young people to become freelancers. The prospect is certainly encouraging, but there is a small hiccup. The term “freelancing training” is dated, not because freelancing is disappearing as a profession—far from it. Thousands of Bangladeshis, like their counterparts across the globe, continue to earn a living by working for overseas clients, and the internet has opened up opportunities that barely existed a generation ago.

So, what’s the problem? It is the word itself. “Freelancing” is not a profession. It is a way of selling work. A freelance software developer, translator, video editor, accountant, architect, cybersecurity specialist, and marketing consultant may have almost nothing in common except that they work independently. Training someone “in freelancing” is therefore a little like training someone “in employment”.

Yet, the Department of Youth Development has proposed spending a huge sum of taxpayers’ money to train over 50,000 young people in freelancing, across 13 metropolitan cities between July 2026 and June 2031. The project expects 60-70 percent of participants to enter income-generating work following training. However, it falls short of answering what exactly these young people will be able to sell five years from now.

It is an urgent question because the market is changing faster than government training programmes can be designed, approved, tendered and delivered. Generative AI has already begun to alter the economics of work traditionally associated with online freelancing. A 2025 study in Management Science, based on a large global freelancing platform, found a 21 percent decline in job postings for automation-prone writing and coding work relative to manual-intensive jobs within eight months of ChatGPT’s arrival. The arrival of AI image generators wiped out 17 percent of job postings for visual design. The remaining jobs tended to be more complex and better paid.

Another study of freelance demand found an even clearer division. Demand for some skills that generative AI could substitute, including writing and translation, fell by 20-50 percent relative to the expected trend. But demand for machine-learning programming rose by 24 percent, while demand for AI chatbot development almost tripled. Even software development, once considered one of the safest career recommendations for an ambitious young person, is changing. “Vibe coding”—building software increasingly through natural language instructions to AI rather than manually writing every line—is an imperfect but revealing sign of where things are heading. The barriers to producing basic websites, applications, graphics, marketing copy, and other digital goods are collapsing.

In Stack Overflow’s 2025 global developer survey, 84 percent of respondents said they were using or planned to use AI tools in software development, while 51 percent of professional developers were already using them daily. At the same time, more developers distrusted the accuracy of AI output than trusted it. This apparent contradiction tells us what valuable skill actually is. It is no longer simply knowing how to produce something. Increasingly, value lies in knowing what should be produced, how to direct the technology producing it, whether the output is correct, and how it solves a real problem.

Judgement, domain expertise, communication, creativity, verification, and the ability to learn new tools have quickly become more important. The World Economic Forum estimates that 39 percent of workers’ key skills will change by 2030. AI and big data, cybersecurity and technological literacy are among the fastest-growing skills, but so are creative thinking, analytical thinking, resilience, flexibility, and lifelong learning.

There is another reason for caution. This is not Bangladesh’s first rodeo with “freelancer training”. The Department of Youth Development launched a Tk 47.5 crore freelancing project covering 16 districts in 2022. This was followed by a roughly Tk 300 crore programme targeting 28, 800 young people across 48 districts. The department’s own project listing shows that the latter programme runs until June 2028.

The earlier pilot produced encouraging headline figures: officials said 64 percent of those trained had secured freelance work. But before committing another huge budget to training, policymakers should ask harder questions. How many were still earning after six months, one year and two years? What was their median income? How much did it grow? What percentage secured consistent income, as opposed to taking on occasional short-term gigs? Which skills generated the highest earnings? How many moved from low-value gigs into higher-value professional work? And most importantly: how did their outcomes compare with similar young people who did not receive the training?

There is also something uncomfortable in the notion that Bangladesh’s low cost of living gives its freelancers an advantage because they can charge international clients less—an argument made in support of the new proposal. Bangladesh should not aspire to become the cheapest pool of digital labour on the internet. That is precisely the part of the market most vulnerable to automation and global competition. If a client’s primary concern is obtaining a logo, a generic article, basic data entry, or a simple website as cheaply as possible, AI increasingly offers an alternative with a marginal cost that approaches zero.

So, what should Tk 683 crore do instead? It should help young people acquire deep, transferable capabilities. That could mean data analysis, cybersecurity, AI-assisted software development, cloud systems, specialised design, financial modelling, industrial technology, digital commerce, or other fields with demonstrable demand. English communication, client negotiation, and portfolio building should sit alongside those capabilities rather than masquerading as the capability itself.

Training should also be modular and continuously updated. A curriculum written in 2026 should not remain intact in 2028. Providers should be able—and required—to replace obsolete modules as technology and employer demand change. The government should also stop treating an account in freelance marketplaces like Fiverr or Upwork as the ultimate destination of digital employment. A capable Bangladeshi professional might freelance, join a local technology company, work remotely for an overseas company, build a small agency, create a software product, or move between all of these during a career.

Bangladesh absolutely should spend public money preparing its young population for the global digital economy. There may be few investments more important. But the objective should not be to manufacture 50,000 or so “freelancers”. It should be to produce 50,000 people capable of learning whatever comes next.

Technology will make some of today’s skills more valuable and others nearly worthless. New professions will appear. Old job descriptions will blur. AI will write code, generate images, analyse documents, and perform tasks we currently pay people to do. Humans, however, will continue to use those systems to create wonders impossible by command-following machines only. The best investment, therefore, is not training young Bangladeshis for a particular marketplace. It is teaching them how to survive when the marketplace changes.

Zarif Faiaz is editor of the Tech & Startup section at The Daily Star and a research fellow at Tech Global Institute. He can be reached at [email protected].

Views expressed in this article are the author's own. 

Follow The Daily Star Opinion on Facebook for the latest opinions, commentaries, and analyses by experts and professionals. To contribute your article or letter to The Daily Star Opinion, see our guidelines for submission.