[🇧🇩] Artificial Intelligence-----It's challenges and Prospects in Bangladesh

[Saif]

How AI became an authority in our classrooms

As AI rapidly enters education, it risks shifting from a supportive tool to an unquestioned classroom authority.

www.thedailystar.net

How AI became an authority in our classrooms

Shamresh Saha

A few months ago, something small happened at home that stayed with me.

My young son had recently discovered what he calls “Gemini Aunty” on my phone. One morning, when neither my wife nor I could persuade him to wake up for school, I tried something different. I asked the AI to speak to him.

A calm voice responded in Bangla, gently telling him that good children wake up on time and prepare for school.

He listened immediately.

What struck me was not just that it worked, but how quickly authority shifted. A child who resisted his parents responded to a machine without hesitation.

That moment raised a question that extends far beyond the home: when AI speaks, what makes us listen?

Across classrooms in Bangladesh, AI is already becoming part of everyday educational practice. Teachers are using it to prepare lessons, generate materials, and manage workloads more efficiently. For many, it is proving to be a valuable support.

But alongside this growing use, a quieter pattern is also emerging.

In Bangladesh, discussions around AI in education often focus on access and adoption. But readiness is not only about whether tools are available. It is about whether educators and institutions are prepared to engage with these tools critically and contextually.

Some educators engage with AI thoughtfully. They adapt its outputs, question its suggestions, and reshape it according to their students’ needs. For them, AI becomes a starting point for thinking.

Others use it differently. Content is generated and used with minimal reflection, treated as ready-made answers rather than something to be examined or questioned. In these cases, AI begins to take on a different role, not just as a tool, but as a source of authority.

This distinction matters.

AI does not simply provide information. It presents it in ways that feel immediate, fluent, and convincing. Its responses often sound complete, even when they are partial or contextually limited. This can create a subtle shift in how knowledge is received. Instead of asking whether something is accurate or appropriate, we may begin by assuming that it is.

In conversations with educators across Bangladesh, this shift is becoming visible. Some teachers describe AI as an essential support that enhances their work. Others express concern that it is increasingly being treated as a “magic solution,” producing answers without requiring much interpretation or adaptation.

This raises a deeper question about readiness.

In Bangladesh, discussions around AI in education often focus on access and adoption. But readiness is not only about whether tools are available. It is about whether educators and institutions are prepared to engage with these tools critically and contextually.

Without that readiness, AI risks being used in ways that reduce professional judgment rather than strengthen it. Lessons may become less responsive to local contexts. Students may rely on answers without questioning how they were generated. Over time, this can shape how learning itself is experienced.

At the same time, avoiding AI is not a realistic option. It is already embedded in how information is accessed and shared.

The challenge, then, is not whether AI should be present in education, but how it is approached.

AI can support teaching and learning in meaningful ways. It can reduce administrative burdens, provide new resources, and open new avenues for exploring ideas. But this potential depends on how it is used.

Visual: Star

It requires seeing AI not as an unquestioned authority, but as something to be engaged with. It requires asking where its responses come from, what they leave out, and how they apply to specific contexts. It also requires maintaining the habit of questioning, even when answers are readily available.

This places responsibility not only on individual teachers, but on the system as a whole. Educators need opportunities to build the skills required to work with AI thoughtfully. Institutions need to create space for reflection, not just adoption. And policy conversations need to move beyond whether AI should be used, towards how it shapes thinking and learning.

AI is already speaking in our classrooms, our homes, and our daily work.

The real question is whether we are still thinking.

Shamresh Saha is a senior manager at the British Council. This article was developed through the #NextGenEdu Learning Cohort, a platform for reflection and dialogue on AI and education in Bangladesh.​

 

[Saif]

AI for an efficient supply chain

In the realm of modern agricultural management, the role of a robust and efficiently managed supply chain in maintaining price stability cannot be overstated. From ensuring the timely movement of goods to prevent hoarding and market manipulation, an effective supply chain serves as the backbone of

thefinancialexpress.com.bd

AI for an efficient supply chain

FE

Published :
Jun 07, 2026 23:03
Updated :
Jun 07, 2026 23:03

In the realm of modern agricultural management, the role of a robust and efficiently managed supply chain in maintaining price stability cannot be overstated. From ensuring the timely movement of goods to prevent hoarding and market manipulation, an effective supply chain serves as the backbone of a well-functioning market. As seen in developed countries, well-integrated supply chains enable agricultural products to move directly from farms to consumers, improving market efficiency while ensuring fair prices for both producers and buyers. In Bangladesh, however, such an integrated system remains largely absent. Given the highly perishable nature of agricultural produce, inadequate storage, logistics and technological infrastructure continue to impede a smooth flow of goods from farms to markets. The resulting inefficiencies not only increase post-harvest losses but also widen the gap between farm-gate and retail prices. Dominance of multiple layers of intermediaries further compounds the problem, often depriving farmers of fair prices while forcing consumers to pay inflated prices.

Against this backdrop, the government's recent proposal to deploy an artificial intelligence (AI)-driven system to monitor supply chains and narrow the gap between farmers and consumers raises cautious optimism. The ambition to embrace data-driven policymaking is welcome. However, AI must not be viewed as a panacea for the structural weaknesses that afflict the supply chain. As per the proposal drafted by the Trading Corporation of Bangladesh (TCB), a digital platform will be created to integrate data from multiple government agencies, markets, logistics providers, warehouses, meteorological office and international commodity indices. By analysing this vast pool of information, the AI system would forecast prices, identify supply disruptions, detect abnormal market behaviour and facilitate more informed policy decisions.

The proposal's emphasis on connecting farmers directly with wholesalers, retailers, e-commerce platforms and government procurement systems is particularly noteworthy. Reducing unnecessary layers of intermediaries could increase farmers' incomes while lowering costs for consumers. An AI-powered monitoring mechanism could also strengthen the government's capacity to respond proactively to market disruptions. Early warning signals would enable the government to make timely decisions on imports, stock management, open-market sales and targeted subsidies. At the same time, the system could help identify unusual price instability arising from hoarding, supply manipulation or artificial market distortions.

However, technology alone cannot solve governance failure or infrastructure challenges. No matter how sophisticated, AI cannot by itself overcome the structural weaknesses that continue to plague the supply chains. For this initiative to succeed, substantial investment will be needed to develop cold-storage facilities across major agricultural hubs and improve transportation networks. Stronger market oversight will also be essential to curb the dominance of syndicates, hoarding and rent-seeking along transport routes. AI can identify where the supply chain is bleeding, but it cannot heal the wound on its own. Overcoming these challenges will ultimately require human intervention and political will. At the same time, the limitations of AI must not be overlooked. Algorithms can produce flawed analyses and inaccurate forecasts. The most effective approach, therefore, would be to combine the speed and analytical power of machines with human judgment and oversight.​

 

[Saif]

Can Bangladesh’s copyright law keep up with AI?

Artificial intelligence is rapidly reshaping the digital world, and Bangladesh is increasingly becoming part of that transformation.

www.thedailystar.net

Can Bangladesh’s copyright law keep up with AI?

Samiul Huq

As generative AI tools are more widely used to create content, a closer inspection of Bangladesh's copyright law has become imperative. FILE PHOTO ILLUSTRATION: REUTERS

Artificial intelligence is rapidly reshaping the digital world, and Bangladesh is increasingly becoming part of that transformation. Generative AI tools such as ChatGPT, Gemini, and Midjourney are now widely used to create written, visual, and software-based content with minimal human involvement. While these technologies offer significant opportunities, they also challenge the traditional foundations of copyright law, which has historically been built to safeguard human labour, creativity, and originality.

Although Bangladesh’s Copyright Act, 2023 modernised several aspects of digital copyright protection, it remains largely silent on the growing problem of AI-generated work and the resulting questions concerning authorship, originality, and accountability.

Modern copyright law is fundamentally built upon the assumption that creative works originate from human intellectual activity. International copyright frameworks such as the Berne Convention for the Protection of Literary and Artistic Works and the TRIPS Agreement implicitly assume the existence of human authorship, without acknowledging mechanical or automated authorship.

J.A.L. Sterling, a significant figure in the advocacy and teaching of copyright law and policy, explained that one of the major philosophical foundations of copyright originates from English philosopher John Locke’s labour theory, wherein individuals acquire property rights over the products created through their own labour, skill, and intellectual effort. Sterling added that copyright law also developed through personality-based theories associated with philosopher Immanuel Kant, where creative works are viewed as extensions of the author’s personality and identity.

AI-generated works destabilise this traditional understanding of “author-work” relationship, according to Paul Goold, because copyright law historically depends upon expressive content being traceable to identifiable human intellectual activity. Goold further argues that modern AI increasingly exposes the conceptual weakness of artificially assigning authorship where meaningful human creativity may not genuinely exist.

Additionally, generative AI creates structural instability within copyright law because machines are now capable of semantic and stylistic imitation at levels previously associated only with human creativity. The concern is therefore no longer limited to copying. AI systems increasingly imitate style, tone, structure, and expressive identity itself.

A recent investigation by fact-checking organisation Dismislab found that two widely used AI models—Gemini and Grok—generated modified NID images using samples and prompts without flagging any concerns about sensitive personal data. In Bangladesh’s current copyright framework, which is still fundamentally designed to assume human authorship and that still struggles to accommodate autonomous AI-generated content, such examples of AI reproducing highly realistic documents should raise concern. In a 2025 paper on the copyright paradigms in the age of AI, the authors observed that concepts such as originality and authorship become increasingly uncertain when expressive works are generated through systems capable of operating with minimal human intellectual contribution.

Therefore, artificial intelligence now requires the law to confront an uncomfortable question: if expressive works can increasingly be generated without meaningful human creativity, what exactly is copyright law protecting anymore?

What Bangladesh needs is not an overly complicated or technologically rigid copyright regime for AI-generated works. Instead, it requires a practical and human-centred framework capable of preserving the connection between copyright and genuine human creativity. Instead of requiring courts or authorities to scientifically determine whether every work is human-created or AI-generated, the law could gradually require creators, publishers, commercial users, or applicants for copyright registration to disclose whether generative AI tools were substantially used in producing the work. The legal focus should then shift towards assessing whether meaningful human intellectual contribution remained dominant in the final output.

Such an approach would also preserve flexibility, as AI technology continues evolving rapidly. Courts could then assess disputes case-by-case by examining factors such as level of human creativity, selection, arrangement, editing, judgment, and intellectual control over the final work rather than attempting to determine whether AI was used at all.

At the same time, Bangladesh should avoid granting full traditional copyright protection to purely autonomous AI-generated outputs where meaningful human creativity is effectively absent. Doing so may gradually weaken the philosophical foundations of copyright law itself by protecting machine-generated production in the same way as human intellectual creativity. Instead, Bangladesh may eventually consider a limited sui generis framework for certain autonomous AI-generated outputs. The term “sui generis” simply means “of its own kind” or a special legal category created outside ordinary copyright law. In intellectual property law, sui generis protection is sometimes used where existing legal categories cannot comfortably address new technological or commercial realities.

Artificial intelligence is forcing copyright law to be reconsidered in terms of the meaning of creativity, originality, and authorship. Bangladesh now has an important opportunity to develop a balanced legal framework that encourages innovation while still preserving the human foundations upon which copyright law has historically depended.

Samiul Huq is a doctoral researcher in law at City St George’s, University of London, UK.​

 

[Saif]

Why AI cannot be left unsupervised

In November 2021, all 193 UNESCO member states unanimously adopted the landmark recommendation on the ethics of Artificial Intelligence (AI).

www.thedailystar.net

Why AI cannot be left unsupervised

Md Mazhar Uddin Bhuiyan

In November 2021, all 193 UNESCO member states unanimously adopted the landmark recommendation on the ethics of Artificial Intelligence (AI). This was the world’s first global framework for ensuring that AI remains transparent, accountable, fair, and subject to meaningful human oversight. Yet despite these efforts, researchers continue to find that many AI systems reproduce racial, gender and social prejudices embedded in the data on which they are trained. A 2024 study found that generative AI tools can amplify stereotypes about race and gender, sometimes producing results even more biased than those found in society itself.

In December 2022, UC Berkeley researcher Steven Piantadosi publicly demonstrated that when an early version of ChatGPT was asked to write a Python function determining whether someone would be a good scientist based on race and gender, it returned “true” only for a white male and “false” for everyone else. In a related prompt, when the model was asked to write code deciding whether a person should be tortured based on country of origin, it produced a function flagging people from specific nations, such as North Korea, Syria, and Iran.

Most people would immediately recognise such outputs as morally unacceptable. But these examples reveal something deeper and far more troubling. Artificial intelligence has no inherent understanding of justice, equality, or human dignity. It learns patterns. If the data it consumes contains prejudice, the machine can reproduce prejudice. If society’s biases are embedded in its training materials, they can emerge in its answers.

For decades, the internet has served as the world’s largest repository of human knowledge. It has also been a repository of human ignorance, discrimination, misinformation, conspiracy theories, and hate. Modern AI systems are trained on enormous quantities of online content. While developers try to implement filters and safeguards, no filtering system is perfect. As a result, AI models may absorb patterns that mirror historical inequalities and stereotypes. Researchers at the U.S. National Institute of Standards and Technology (NIST) have warned that AI bias originates not only from data but also from broader societal structures and human decision-making processes.

Researchers have found that many AI image-generation tools do not represent people equally. Women and people of colour are often shown less frequently in high-status professions and more frequently in lower-paying jobs. A 2023 Bloomberg investigation analysing more than 5,000 images generated by Stable Diffusion found that over 80 percent of people shown in high-paying professions, such as CEOs and lawyers, had lighter skin tones.

Artificial intelligence has no inherent understanding of justice, equality, or human dignity. It learns patterns. If the data it consumes contains prejudice, the machine can reproduce prejudice. If society’s biases are embedded in its training materials, they can emerge in its answers.

Women were also significantly underrepresented. For example, the AI portrayed women as judges only 3 percent of the time, even though women make up about one-third of judges in the United States. These findings suggest that AI systems can reinforce and even amplify existing social stereotypes. UNESCO found similar gender bias in AI models. A 2024 UNESCO study of GPT-2, GPT-3.5, and Llama 2 showed that these models linked men more closely with leadership, science, technology, and career-related words, while linking women more closely with domestic roles. One model found that women were described in domestic roles four times more often than men.

Meanwhile, governments and companies are increasingly using AI in hiring, policing, welfare, education, and healthcare. In the United Kingdom, a February 2024 government fairness review found that an AI welfare fraud system used for Universal Credit Advances showed statistically significant differences across age, disability, marital status, and nationality. However, the Department for Work and Pensions said there were “no immediate concerns of unfair treatment.”

These cases expose a dangerous assumption that often accompanies technological innovation, that algorithms are neutral. They are not. An algorithm trained on biased information can make biased recommendations. A machine learning model optimised for historical outcomes can reproduce historical injustices. A generative AI system trained on discriminatory content can generate discriminatory responses.

Now a new threat is emerging. As AI-generated content floods the internet, future AI models may increasingly be trained on content created by previous AI systems. Researchers have begun warning about a feedback loop in which synthetic data contaminates future training datasets. In simple terms, machines may begin learning from machines.

A 2024 Nature study by Oxford and Cambridge researchers found that AI models can suffer “model collapse” when they are repeatedly trained on AI-generated content. Over time, they first lose rare patterns and minority data. Some analysts warn that AI could generate up to 90 percent of online content within a few years.

What is the potential danger in this ecosystem? Imagine photocopying a document thousands of times. Each copy introduces tiny imperfections. Eventually, the original image becomes distorted beyond recognition. Something similar may occur with AI-generated knowledge. If biased, inaccurate, or fabricated AI content spreads across websites, blogs, forums, and social media, future systems may absorb and reinforce those distortions.

Importantly, human oversight is not a sign of technological weakness. It is a recognition of technological reality. Even the most advanced AI systems lack moral judgment. They do not understand fairness. They do not comprehend historical injustice. They cannot independently determine whether an output is ethically acceptable. Humans can. Human review is not an obstacle to innovation; it is a prerequisite for trustworthy innovation.

The result could be a more binary world, one where nuance disappears, stereotypes harden, and automated systems increasingly categorise people into simplistic groups. This is precisely why human oversight must remain at the centre of AI development.

Thus, UNESCO’s recommendation to incorporate human oversight at every stage of AI development should be supported as a core principle of responsible AI governance. Human-in-the-loop systems offer one of the most effective safeguards against algorithmic harm. Rather than allowing AI systems to operate autonomously, human reviewers can monitor both the information AI systems consume and the outputs they generate. Experts can identify discriminatory patterns, verify factual accuracy, challenge questionable recommendations and intervene when models produce harmful content.

Importantly, human oversight is not a sign of technological weakness. It is a recognition of technological reality. Even the most advanced AI systems lack moral judgment. They do not understand fairness. They do not comprehend historical injustice. They cannot independently determine whether an output is ethically acceptable. Humans can. Human review is not an obstacle to innovation; it is a prerequisite for trustworthy innovation.

Governments should therefore require independent audits of high-risk AI systems. Companies should maintain diverse human review teams capable of identifying cultural, racial, gender, and political biases. Educational institutions should teach algorithmic literacy so citizens understand how AI systems influence their lives. Most importantly, developers should be required to document training sources and demonstrate how bias testing is conducted before deployment.

History teaches us that every transformative technology requires guardrails. Railways needed safety standards. Pharmaceuticals require clinical trials. Aviation demanded rigorous oversight. Artificial intelligence should be no different.

The question is not whether AI will shape our future; it already is doing so. The question is whether that future will be shaped solely by patterns extracted from the past or guided by human values capable of correcting them. However, it is good that OpenAI, Anthropic, and Google are regularly updating their policies to make information more reliable.

In conclusion, machines can process information faster than any human being. But they cannot decide what kind of society we want to build. That responsibility remains ours. For that reason, every AI system that affects human lives should contain something no algorithm can replace, that is, a human being in the loop.

Md Mazhar Uddin Bhuiyan is an Oxford-Felix scholar and Master of Public Policy candidate at the University of Oxford.​

 

[Saif]

AI and loss of linguistic creativity

WHILE artificial intelligence has much to offer to language, communication, education and society, it may also erode linguistic creativity. This may be in relation to English in particular, the global lingua...

www.newagebd.net

AI and loss of linguistic creativity

WHILE artificial intelligence has much to offer to language, communication, education and society, it may also erode linguistic creativity. This may be in relation to English in particular, the global lingua franca, which provides almost 90 per cent of the training data for generative AI systems.

Specifically, our concern is the linguistic creativity by non-native speakers of English. They are a clear demographic majority as there are three times more such speakers of English than its native speakers globally. Practically, more communication involving English now takes place within the former community than in the latter. Nevertheless, non-native speaker creativity in English has been slow to win recognition. For a long time, English language learners have been viewed as shadows of native speakers. They are supposed to pursue native-speaker competence as the only legitimate learning goal. Their learning journey starts from their first/native language, and they travel up to the almost unreachable ideal of native-speaker ability. For most of them, the journey ends long before they reach the destination. A few who are lingua-savvy may climb the height, but they may still not be considered native speakers due to, among other factors, the colour of their skin. Fortunately, things have been changing with non-native speakers of English outnumbering native speakers, but the pace has been much slower than expected.

It is the linguistic creativity of these learners and speakers of English that is my point. They are located at all stages of their language learning and using trajectory. Certainly, they don’t make one homogenous group. The contexts and circumstances of their English learning are different; different is also the extent of the use of English in their everyday life.

Nonetheless, many can be creative in English, with all their mistakes, idiosyncrasies and imperfections. As AI comes to dominate our language life and communication, this creativity is likely to disappear.

How may that happen?

Creativity in language can’t be associated only with native intuition or linguistic expertise. It also arises when speakers try to communicate with limited language skills. Linguistic creativity can emerge even from errors or imperfect learning when there is an urge for self-expression. For example, the innovative expression ‘Long time, no see’ is grammatically incorrect, yet its innovative expression of emotion is unique and undeniable. The origin of this expression is debated but it is widely viewed as an example of pidgin English which originated in a Chinese-English contact situation. The innovator of this creativity did not have perfect English. However, their limited linguistic ability to survive the communicative imperative gifted a profound expression to the English language.

This innovation is now considered part of mainstream English. Its grammatical oddity may not invite language policing or an orthodox English teacher using their red pen. It’s a classic example of how something that is an error at origin can travel a long way to be accepted as a legitimate expression.

Alas! with AI, speakers of pidgin English may not venture into such creative expressions. As people rely more on AI for writing and other communication, the everyday creativity through linguistic struggle may vanish. It is critical to appreciate this hidden loss of human creativity as native speaker-controlled AI dominates communication, life and society.

Another example is a historical letter which is currently on display at the Railway Museum in New Delhi. As reproduced below, it was written by one Okhil Chandra Sen to the divisional railway office in 1909 during British colonial rule in India. The ungrammatical text authored by a half-educated Indian man marked his complaint, also pointing to the limits of his linguistic ability.

‘I am arrive by passenger train Ahmedpur station and my belly is too much swelling with jackfruit. I am therefore went to privy. Just I doing the nuisance that guard making whistle blow for train to go off and I am running with ‘lotah’ in one hand and ‘dhoti’ in the next when I am fall over and expose all my shocking toman and female women on plateform. I am got leaved at Ahmedpur station. This too much bad, if passenger go to make dung that dam guard not wait train five minutes for him. I am therefore pray your honour to make big fine on that guard for public sake. Otherwise I am making big report! to papers.’

If we ask ChatGPT or another AI tool, it can do a good translation of this ‘broken’ English into British or American variety. But the writer’s meaning can be deciphered even from the non-standard text. His communicative intention is not lost in ungrammaticality or linguistic border crossing. Trains in those days didn’t have toilets in India. As the train that carried Mr Sen stopped at the noted station, he got off to respond to the call of nature in the open space. However, the train left before he was able to return to board it. As he was running to catch the train managing his dhoti and water pot, he ended up exposing himself to the surrounding world.

People may judge his English in any way they like. The judgment may depend on whether one is a linguistic puritan or a liberal. To me, this is real English in the life circumstances of this man and his Indian social surroundings. Expecting him to be able to write immaculate British or American English or even a more polished Indian variety will be unreasonable.

Our language reflects our life and situations. It can’t be too distant from our social existence. However, what is noteworthy is the urge to communicate, despite the limited ability. Despite all oddities, I am happy to consider it an innovative text that successfully communicated the message. I can’t but appreciate the linguistic graciousness of the audience who took the message and ignored the grammar.

We shouldn’t forget that this letter of complaint led to introducing toilets in Indian trains. As speakers of standard English, we may police the language, but we can’t undermine the impact it had. It is this achievement that has made the letter an artefact to be displayed in a museum.

Unfortunately, with AI tools in our hand, we may not see such creativity or innovation in language use. We may not even exert our intellectual, cognitive or creative labour. We will ask AI and it will do what we want done. We may even stop struggling with English.

Instead of experimenting on their own, nonnative speakers may use AI to produce correct language at the expense of their creative potential. While this makes communication easier, it may also eliminate their creative urge and homogenise the linguistic diversity that non-native speakers bring.

This potential loss of linguistic creativity will be a loss of humanity, human culture and civilisation.

Obaidul Hamid is an associate professor at the University of Queensland in Australia. He researches language, education and society in the developing world. He is a co‑editor of Current Issues in Language Planning.​