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

[Saif]

Bangladesh must seize AI opportunities amidst global rivalry

The rivalry between America and China regarding artificial intelligence (AI) is growing

www.thedailystar.net

Bangladesh must seize AI opportunities amidst global rivalry

PHOTO: REUTERS

The rivalry between America and China regarding artificial intelligence (AI) is growing, with both nations significantly investing in the technology. Historically, AI development has been dominated by America, but China is closing in at a rapid pace. One such example is the emergence of DeepSeek, a Chinese AI startup that has created AI models capable of competing with those from the Silicon Valley. DeepSeek also made their product accessible through an open-source model. This competition has profound implications for technology, economies, and politics worldwide.

Bangladesh—a developing economy, must strategically position itself in the AI race to fully harness its potential for technological transformation, economic growth, and innovation.

The competition between America and China

The US holds a dominant position in AI development, commercialisation, and research worldwide. Leading US-based tech companies—such as OpenAI, Google DeepMind, and Meta have pioneered AI development, creating models like ChatGPT, Gemini, and Llama. Recognising AI's strategic value, the US government has invested significantly in AI-driven national security and economic competition.

China got advanced AI through state-backed initiatives, focusing on industrialisation and national security applications. Tech giants like Tencent, Alibaba, and Baidu have built powerful AI models—such as ERNIE and Qwen 2.5 Max. DeepSeek challenges Silicon Valley's dominance with lower costs. As competition intensifies, countries like Bangladesh must navigate the evolving AI landscape.

Bangladesh's current AI landscape

In an effort to embrace AI, Bangladesh introduced the National Artificial Intelligence Policy 2024, which aims to guide AI development through research, education, and ethical implementation. While the policy lays a foundation for AI adoption, significant gaps remain in infrastructure, vision, and execution. The private sector and startups have shown interest in AI, but a lack of funding and computing resources remains a major hurdle.

AI is gradually being integrated into key sectors of Bangladesh's economy. In healthcare, AI-driven diagnostics and predictive analytics are being explored to improve patient outcomes. The agricultural sector is benefiting from AI-powered precision farming and pest control strategies. Financial institutions are leveraging AI for fraud detection and customer analytics, while the manufacturing industry is experimenting with automation, albeit at a relatively low scale. Despite these advancements, businesses are still having to rely on foreign AI models, which increases operational costs and limits data sovereignty.

Challenges hindering Bangladesh's AI growth

One of the biggest challenges Bangladesh faces is that without domestic AI capabilities, Bangladesh remains vulnerable to external market fluctuations and pricing structures dictated by foreign providers.

Another major hurdle is the lack of infrastructure. Bangladesh currently has no large-scale AI data centres, forcing researchers and companies to rely on expensive cloud computing services from AWS, Google Cloud, and Microsoft Azure. The absence of high-performance computing (HPC) infrastructure slows down AI research and limits local innovation.

Additionally, Bangladesh faces a shortage of skilled AI professionals. The country's top talents tend to migrate abroad in search of better opportunities, leading to brain drain. While local universities offer AI-related courses, AI research has yet to be fully integrated into their core curriculum. Without a strong pipeline of AI-related talent, Bangladesh will struggle to build a self-sufficient AI ecosystem.

Opportunities for Bangladesh amidst the AI rivalry

Despite these challenges, Bangladesh has several opportunities to establish itself as a competitive player in this sector. One promising approach is leveraging open-source AI models such as DeepSeek's V3 and R1, Meta's Llama, Mistral, and Falcon and others. Open-source AI models offer greater flexibility, cost efficiency, and data sovereignty, reducing the country's reliance on expensive proprietary systems from global AI leaders.

Another crucial step for Bangladesh is developing its own AI foundation model trained in Bangla and regional datasets. By creating localised AI models, Bangladesh can address the needs of local businesses, government agencies, and educational institutions. A home-grown localised foundation model would enable the country to build AI applications tailored to its unique economic and linguistic landscape, reducing dependence on foreign technology.

To make these initiatives viable, investing in local AI data centres is essential. Establishing AI infrastructure would lower computing costs, making AI adoption more accessible for businesses and startups. Local AI data centres could also attract companies seeking affordable, regionally hosted AI solutions, creating new economic opportunities.

Policy recommendations for strengthening Bangladesh's AI ecosystem

To accelerate AI growth, the government must increase funding for AI research and introduce AI-specific grants and tax incentives for startups. By investing in AI innovation, Bangladesh can stimulate local development and attract foreign investment.

Educational institutions must play a crucial role in this transformation by collaborating with industry leaders to develop AI-focused research programmes. The government should also establish dedicated AI research institutions that focus on cost-effective, scalable AI solutions tailored to Bangladesh's needs.

Public-private partnerships will be key to advancing AI development. The government should foster collaboration between academia, businesses, and policymakers to create a cohesive AI ecosystem. Regulatory sandboxes should be introduced, allowing AI startups to test their models with minimal bureaucratic restrictions.

Another critical step is implementing data localisation laws to ensure that sensitive data is stored within Bangladesh, reducing security risks and enhancing data sovereignty. Additionally, strong AI governance frameworks should be developed to prevent unethical AI use and ensure responsible AI deployment.

Despite lacking AI infrastructure and skilled professionals, Bangladesh can establish a unique role in the global AI arena. This will be a technological leap with economic imperative for future competitiveness.

Rashedul Huq is research associate at North South University.​

 

[Saif]

AI is ‘biggest shift of our lifetimes’

Artificial intelligence (AI) is a “fundamental rewiring of technology” that will act as an “accelerant of human ingenuity”, Google boss Sundar Pichai will say Monday at a global AI summit in Paris.

www.thedailystar.net

AI is ‘biggest shift of our lifetimes’

Artificial intelligence (AI) is a "fundamental rewiring of technology" that will act as an "accelerant of human ingenuity", Google boss Sundar Pichai will say Monday at a global AI summit in Paris.

"We're still in the early days of the AI platform shift, and yet we know it will be the biggest of our lifetimes," Pichai was to say according to excerpts from his speech to global leaders and tech industry chiefs seen by AFP.

The Google boss will highlight imminent applications for AI technology such as detecting the emergence of wildfires in satellite images, as well as announce a partnership for detecting and treating cancer with the French capital's Institut Curie.

Demis Hassabis, the head of Google's DeepMind AI research lab, also hailed the technology's potential at an event on Sunday in Google's Paris offices.

"Material science, mathematics, fusion, there is almost no area of science that won't benefit from these AI tools," the Nobel chemistry laureate said.

"With AI, we have the chance to democratise access (to a new technology) from the start, and to ensure that the digital divide doesn't become an AI divide," Pichai was to say Tuesday.

The concept of a "digital divide" refers to the gap between people with or without easy access to and familiarity with the internet and modern communications technologies.​

 

[Saif]

Paris summit grapples with real-world impact of AI

Artificial intelligence (AI) risks reinforcing gender disparities rather than immediate widespread job destruction, the International Labour Organization chief said yesterday as a summit of national leaders and tech bosses opened in Paris.

www.thedailystar.net

Paris summit grapples with real-world impact of AI

Artificial intelligence (AI) risks reinforcing gender disparities rather than immediate widespread job destruction, the International Labour Organization chief said yesterday as a summit of national leaders and tech bosses opened in Paris.

Co-hosted by French President Emmanuel Macron and Indian Prime Minister Narendra Modi, the Paris summit aims to lay the groundwork for governing the nascent sector, as global powers race to play leading roles in the fast-developing technology.

At present, AI is mostly replacing humans in clerical jobs disproportionately held by women, International Labour Organization (ILO) head Gilbert Houngbo told an audience in the French capital's opulent Grand Palais.

"The category of jobs that will be automated and hence be lost could contribute to deepen the gender gap" in pay between men and women, he added.

Looking at the wider picture, "we're not too much worried (about) a massive job loss" despite widespread fears Houngbo added.

While "there will be millions of jobs that will be lost", many more will be created thanks to AI, he predicted based on current "quite positive" trends.

Nevertheless, "there is a risk of those new jobs being paid less and sometimes with much less protection" for employees, as has happened with much "gig economy" work, Houngbo said.

Political leaders, including US Vice President JD Vance and Chinese Vice Premier Zhang Guoqing, are set to rub shoulders with the likes of OpenAI boss Sam Altman and Google chief Sundar Pichai.

A largely suit-wearing crowd of men and women speaking languages from all over the world gathered under the glass-and-steel dome of the great hall, built for the 1900 Universal Exhibition and now decked out with screens and geodesic domes.- 'Immense hopes, exaggerated fear' -

Two years on from the emergence of OpenAI's ChatGPT chatbot, "artificial intelligence fuels both immense hopes and, at times, exaggerated fear," Macron's AI envoy Anne Bouverot told guests as she opened the summit.

She promised a "turning point" that would bring more countries on board with AI, as well as more "sustainable development" of the resource- and energy-hungry technology.

Macron had on Sunday trumpeted the economic benefits of artificial intelligence, saying 109 billion euros ($113 billion) would be invested in French AI in the coming years.

That was "the equivalent for France of what the US has announced with 'Stargate'," the $500-billion US programme led by ChatGPT maker OpenAI, he added.

The technical challenges and price of entry for nations hoping to keep abreast in the AI race have become clearer in recent week​

 

[Saif]

Businesses riding the AI wave

About seven years ago, representatives of bKash would visit the company’s agent points and merchant outlets across Bangladesh to ensure that its point-of-sale and promotional materials, such as banners, posters and stickers, are properly displayed.

www.thedailystar.net

Businesses riding the AI wave

Illustration: Biplob Chakroborty

About seven years ago, representatives of bKash would visit the company's agent points and merchant outlets across Bangladesh to ensure that its point-of-sale and promotional materials, such as banners, posters and stickers, are properly displayed.

These materials were essential for conveying promotional offers and other marketing messages from the country's leading mobile financial service provider.

But in 2018, bKash partnered with Intelligent Machines, an artificial intelligence (AI) company, to complete such tasks using computer vision, which is a field of AI that enables machines to analyse and interpret visual data.

Called Biponon, the service was then introduced across all of its roughly eight lakh outlets, subsequently streamlining what was once the cumbersome task of managing trade marketing and sales operations.

The results were astounding as bKash achieved a 320 percent rise in productivity, with the AI tool enabling its merchandisers or representatives to serve 585 outlets compared to 139 previously.

This is just one example of how the wave of AI, which is sweeping the world like a tsunami while also sparking geopolitical tensions, has reached the country's shores.

Following the global trend of embracing AI to reduce human intervention and improve operational efficiency, local companies have started using it for predictive maintenance, inventory management and supply chain optimisation, albeit at a slower pace.

Now, some local AI development companies are driving innovation across various industries through their products. In essence, they are developing AI-powered solutions to address challenges in areas such as business, agriculture, healthcare, finance and education.

Globally, Walmart leverages AI for forecasting demand and inventory optimisation while Amazon uses it for warehouse automation and delivery planning. In the financial sector, AI is being utilised to analyse customer behaviour and detect fraud.

Likewise, some of the leading local companies, including Unilever, Grameenphone and Banglalink, alongside startups like Pathao, and e-commerce platforms like Foodpanda and Daraz, are adopting AI to optimise operations, personalise customer experiences and enhance efficiency.

McKinsey and Company, a US-based multinational strategy and management consulting firm, said AI adoption could contribute about $13 trillion to the global GDP by 2030, with the retail, healthcare and manufacturing sectors being key beneficiaries.

By automating repetitive tasks and enabling real-time decision-making, AI reduces costs and unlocks innovative growth opportunities, reshaping global business practices.

To harness AI's potential, two types of companies have emerged worldwide: those that use AI solutions and those that build AI while some are hybridising by developing tailored AI solutions.

And Bangladesh is no exception to this ongoing transformative trend.

COMPANIES THAT EMBRACE AI

Other than using AI to streamline its point-of-sale materials, bKash also deeply integrated AI into several core business functions, particularly customer service, data management and decision-making.

"For instance, we leverage predictive modelling to gain insights into customer behaviour, which enables us to identify trends and implement adequate interventions to enhance customer experiences and retention," said Azmal Huda, chief product and technology officer of bKash.

"In customer service, intelligent AI-driven chatbots provide instant responses to common queries, streamlining service delivery and reducing operational load," he added.

bKash also leverages AI to enhance the customer experience and optimise service delivery. Its recommendation engine analyses user preferences and behaviour to suggest relevant services, increasing customer engagement and satisfaction.

AI-driven credit scoring models facilitate micro-loan services by accurately assessing creditworthiness, ensuring timely access to credit. An AI-based loyalty engine distributes loyalty points, encouraging long-term engagement and driving revenue growth. Machine learning (ML) streamlines its customer onboarding by verifying identity documents and facial data. Predictive modelling and forecasting further enable bKash to anticipate market trends and adapt its offerings accordingly.

Demand forecasting powered by AI ensures that system capacities are optimised, reducing over-provisioning and underutilisation of resources. Additionally, we employ predictive modelling and anomaly detection techniques to identify and mitigate fraud, Huda said.

"AI will be a key differentiator, enabling companies to extend next-generation financial services, such as credit access and smart saving solutions, to underserved populations, thereby driving financial inclusion and market expansion," he added.

bKash prioritises skill development through knowledge-sharing sessions and workshops, fostering innovation and empowering employees for an AI-driven future.

Banglalink, a leading mobile network operator and digital service provider, has integrated AI deeply into its operations to enhance customer experience and efficiency.

AI powers various aspects, including customer service, product development and internal processes. Generative AI chatbots provide instant and accurate support, significantly improving customer satisfaction.

The AI-driven platform RYZE offers personalised services like CV writing and profile enhancement, with more features underway. Also, the MyBL app curates relevant content, streamlining user experiences.

Besides, AI-driven data analysis enables Banglalink to segment customers, personalise offerings and optimise operations through revenue forecasting, stock predictions and credit scoring. The company also employs P.I. Works' AI-powered solutions with self-organising network (SON) capabilities to enhance network performance while reducing human error.

Looking ahead, Banglalink plans to expand AI adoption in education, health and fintech while training employees to maximise its potential, ensuring continued growth and innovation.

Unilever uses its precision marketing AI tool provided by Intelligent Machines, Fordo, to recommend personalised discounts to 2.3 million customers at about 4,500 kitchen markets all over the country.

The Unilever team has achieved successful conversion rates of 6 percent and above (often as high as 13 percent), up from 1 percent before using AI.

Unilever Bangladesh said it initially focused on software, but has now integrated AI and ML to automate key processes. Furthermore, AI-driven stock replenishment, sales strategies, and route optimisation have enhanced company efficiency. A dedicated analytics team leverages ML for market insights.

UNA, an AI-powered generative platform, supports employees by streamlining tasks, answering queries and providing quick access to essential resources, improving productivity and user experience across various functions.

British American Tobacco Bangladesh (BATB) has taken their Bangla keyword spotting AI, Shobdo, to track and improve 5,540 sales representatives' brand message deliveries at 200,000 outlets. The team has been able to increase the effectiveness of their brand message deliveries from 15 percent to 75 percent using AI.

BATB has also taken their predictive maintenance AI, Jontro, to determine the optimum time to replace the spare parts of the factory's machines.

IDLC Finance and Prime Bank are using a financial statement analysing AI, Dharapat, to process bank statements and the central bank's Credit Information Bureau report, bringing down the critical operations turnaround time from days to minutes.

Grameenphone and Ericsson recently announced a partnership for leveraging AI to enhance telecom services and drive digital transformation in Bangladesh. Their partnership focuses on upgrading Business Support Systems with AI-powered solutions, such as intelligent usage analysis, anomaly detection and automated service management.
Robi has introduced an AI-powered distribution bot to revolutionise its supply chain, addressing challenges in serving over 55 million subscribers across Bangladesh.

Besides, AI solutions have reduced its low-balance scenarios by 25 percent and improved service reliability.

Bangladesh's largest consumer tech platform, Pathao, is also leveraging AI and custom machine learning models to streamline operations and improve user experiences across its ride-hailing, delivery and fintech platforms.

These include calculating accurate fares, estimating trip durations, suggesting destinations and optimising ride pairings. Advanced geocoding ensures efficient parcel delivery routes while AI verifies eKYC for fintech and validates content for quick commerce.

Future developments include automating customer support, enabling conversational product discovery, performing credit scoring and determining the best customer lifecycle interventions.

The manufacturing sector is also embracing AI, with companies like Akij Group adopting predictive maintenance systems to minimise machinery downtime.

Retailers like Daraz Bangladesh and Chaldal rely on AI for personalised customer recommendations and demand forecasting, streamlining inventory management. Even education is evolving: Edtech platforms like Shikho and 10 Minute School use AI to tailor learning experiences, adapting content to individual student performances. These examples highlight how Bangladeshi industries are harnessing AI to solve local problems while boosting productivity.

Walton is transforming home appliances with AI-driven innovations. Its smart refrigerators use an AI-based multiple-speed optimisation algorithm to optimise cooling and reduce energy consumption.

For example, AI-powered air conditioners learn user preferences for personalised comfort.

Meanwhile, the company's IoT-enabled electrical appliances allow users to create customised automation scenes, with the Walton Smart Appliances app enabling remote monitoring and control.

AI-BUILDING COMPANIES

With demand for AI solutions or models on the rise, AI-building companies specialising in developing AI models and solutions for various industries have bourgeoned in Bangladesh in the last few years.

These companies focus on creating custom AI tools, such as ML algorithms, computer vision systems, and natural language processing models.

Youths have taken centre stage here, founding a number of companies focused on creating AI solutions for industries such as telecommunications and security. These AI-building companies help businesses automate processes, optimise operations and improve decision-making, driving digital transformation.

Sigmind, a local startup founded in 2017, is primarily leveraging computer vision, a branch of artificial intelligence that focuses on enabling machines to interpret and analyse visual data from the real world.

It now specialises in advanced AI-driven solutions for vehicle analytics, including vehicle identification, classification, number plate recognition, and traffic and toll management systems.

"Our technology enhances transparency and security in environments such as export processing zones by automating processes like vehicle counting and access control," said Md Abu Anas Ibn Samad, founder and CEO of Sigmind.

For human analytics, Sigmind offers automated attendance systems, real-time entry monitoring, identification and access control. This eliminates the need for manual attendance processes and ensures seamless security through anomaly detection and alerts.

Sigmind has deployed its solutions across seven countries, serving governments and enterprises with a team of 18 professionals.

"We are driving AI adoption to enhance efficiency and foster development. As AI continues to transform industries, the need for upskilling the workforce is crucial," Samad said.

CURRENT STATE OF AI IN BANGLADESH

Bangladesh's position in AI is still nascent compared to global leaders like the US, China and India.

According to experts, while local companies are leveraging AI for operational efficiency and customer engagement, the country lacks a robust AI ecosystem.

Key challenges include limited infrastructure, skill gaps and insufficient investment in research and development.

To compete globally, Bangladesh must prioritise a national AI strategy, invest in education and training, foster public-private partnerships, and encourage local innovation.

"Key challenges for Bangladesh include fragmented academia-industry ties, scarce R&D funding, and policy gaps. Prioritise tripartite collaboration among academia, industry, and policymakers," said Mohammad Mahdee-uz Zaman, an AI policy expert.

He further said academia should develop AI-focused curricula while industries should provide practical training and policymakers should incentivise innovation.

"Strengthen digital infrastructure, invest in AI research centres and promote rural AI literacy. By aligning demographic agility with strategic upskilling, Bangladesh can leverage AI for domestic problem-solving in sectors like agriculture and healthcare, while positioning itself for global market leadership," he added.

Zaman, also the founder of CloudCamp Bangladesh, said with 60 percent of its population under 30, Bangladesh's youth and 1 million freelancers are pivotal to harnessing AI.

So, upskilling this workforce in AI tools (data annotation and ML models) can position the country as a global outsourcing hub for AI-driven tasks.

According to Mohammad Oli Ahad, the founder of Intelligent Machines and a passionate AI practitioner, the world is much like where it was in the early 1990s with the internet when it comes to AI.

"Similar to how, with the exceptions of socioeconomic left-outs, every industry and business now uses the internet, they will be using AI in coming years," said Ahad, who left Intelligent Machines and founded a new startup called rmg.ai, aiming to help transform the country's leading export industry.

"AI will break down every business model and put them back together in new ways," he added.

Rakibul Hassan, one of the country's leading authors on AI and ML, said adopting AI at scale requires strategic best practices, and the government should spearhead a national AI strategy with sector-specific roadmaps emphasising agriculture and healthcare.

Hassan suggested that a dedicated council, comprising experts from academia, industry and policymakers, could guide implementation. Public-private partnerships (PPPs) are equally vital as collaborating with tech firms to develop AI solutions for public services, such as smart traffic systems in Dhaka, would demonstrate tangible benefits.

Meanwhile, investing in digital infrastructure, including high-speed internet and cloud computing, is essential to support startups like Chaldal and Daraz.

"Universities like BUET and all others must integrate AI into STEM curricula to nurture talent," said Hassan, who is also chief technology officer at CTO Link3 Technologies.

Projects in low-risk areas, such as using ML to predict water/power demand in Dhaka, can build confidence in wider deployment, he added.

He also said challenges persist, including limited data quality, skill gaps and resistance to change.

However, Bangladesh's young, tech-savvy population and thriving startup ecosystem offer unique opportunities.

"By prioritising ethical frameworks, infrastructure and collaboration, Bangladesh can position itself as a regional AI hub, using technology to tackle pressing issues like climate resilience, healthcare access and economic inclusion."​

 

[Saif]

AI, data-driven strategies transforming business operations: experts

The daylong Digital Summit held in Dhaka

www.thedailystar.net

AI, data-driven strategies transforming business operations: experts
The daylong Digital Summit held in Dhaka

Artificial Intelligence (AI) and data-driven strategies are transforming modern business operations, experts said today.

AI has become a fundamental tool in solving business challenges, with data analytics, machine learning, and advanced algorithms enabling businesses to anticipate market demand, manage inventory efficiently, and refine marketing strategies for maximum impact, said Zaved Akhtar, CEO of Unilever Bangladesh.

"With the rise of multiple TV channels and the evolution of digital marketing, firms must adapt and shift their perception accordingly," he said.

Akhtar also highlighted AI's capabilities, stating that it can now prepare, simulate, and complete 3D printing in just 30 minutes, whereas it previously took 15-20 days.

Bangladesh can be broadly divided into two regions: the eastern region, which is more industrialised and receives significant remittances—leading to a higher concentration of digital advertising—and the western region, which is predominantly rural, where door-to-door sales are more prevalent, said Akhtar, who is also the president of the Foreign Investors' Chamber of Commerce & Industry.

He was delivering a keynote address at the 11th edition of the Digital Summit, presented by Meghna Group of Industries (MGI) and organised by Bangladesh Brand Forum at InterContinental Dhaka.

The event served as a platform for industry leaders, digital experts, marketing professionals, and business strategists to convene and explore the dynamic and ever-evolving digital landscape.

The summit facilitated in-depth discussions on how businesses and brands can navigate the complexities of digital transformation while staying attuned to shifting consumer behaviours and emerging technological innovations.

Guided by the overarching theme "Digital Renaissance: Innovating for Tomorrow's Consumer," this year's summit analysed the intersection of technology, consumer psychology, and brand engagement.

Discussions underscored how brands must rethink their strategies to forge stronger and more meaningful connections with the next generation of consumers, particularly Gen Z and Gen Alpha.

The event featured a meticulously designed agenda, comprising an exciting lineup of two keynote sessions, four panel discussions, and two insight sessions, all aimed at offering deep insights and practical solutions for digital transformation.

Delivering the opening remarks, Sajid Mahbub, group CEO and executive editor of Bangladesh Brand Forum, emphasised the importance of embracing digital evolution as a necessity rather than an option.

"We are witnessing an era where digital transformation is no longer a distant concept—it is happening all around us, redefining how businesses operate and how brands engage with their audiences," he said.

"The digital renaissance calls for innovation, agility, and a profound understanding of consumer expectations. This summit is designed to equip industry leaders with the knowledge, tools, and strategies to rethink their approaches, foster deeper engagement, and navigate the complexities of today's fast-changing digital terrain."​

 

[Saif]

National AI Policy must address vulnerabilities of artificial intelligence

Bangladesh must ensure its AI system is fair, secure, and beneficial for all.

www.thedailystar.net

National AI Policy must address vulnerabilities of artificial intelligence

File Illustration: Salman Sakib Shahryar

Last year, the Bangladesh government released a draft of the National Artificial Intelligence (AI) Policy, which, given the changing political situation and a lack of public interest in general, did not get enough attention. Despite projecting a lofty vision of catapulting Bangladesh into an era of AI innovation and adoption by harnessing it for the well-being of citizens, economic prosperity, and sustainable development, the draft policy falls short in some aspects, especially compared to the more detailed and exemplary regulatory frameworks such as the Artificial Intelligence Act in the European Union (EU) and AI regulations in China.

One of the vital goals of any national AI policy should be ensuring that AI systems are legally required to be reasonably safe, secure, reliable, and protected against errors and biases. That is precisely one of the areas where our policy could be more comprehensive.

One of the major risks of AI lies in its susceptibility to errors and biases, which raises a wide range of ethical concerns. Unlike conventional computer programmes, AI systems powered by machine learning can learn from data without explicit directives from human operators. While this gives AI transformative potential—such as in the case of operating self-driving cars and creating art or poetry—it also makes them susceptible to errors.

The first type of error comes from the data used to train the AI. It is easier to explain this with a few examples. In 2018, Amazon, the world's largest e-commerce retailer, used an AI recruitment tool that helped check applications. The tool had been trained on checking the CVs submitted to the company in the last 10 years, most of which came from men. Consequently, the algorithm learnt to favour male candidates, and downgrade CVs that included the word "women." After the issue was discovered, Amazon scrapped the tool. Similarly, in 2020, Twitter faced backlash when users discovered that its image-cropping algorithm favoured white faces over black ones. The algorithm was found to have been trained on datasets that did not adequately represent different skin tones, leading to biased decisions that favoured lighter-skinned individuals. Therefore, there is a critical need for diverse and representative datasets when training machine learning models to avoid biases and ensure fair and accurate performance.

The second source of errors is the algorithms themselves. For instance, in 2023, autonomous vehicle manufacturer Cruise had to recall its entire fleet of self-driving cars in the US after a vehicle struck a pedestrian, causing severe injuries. These cars were using a particular class of machine learning—deep learning, which is the technology behind today's most advanced AI systems, but it is not yet advanced enough to anticipate incidents like this. Regardless of the volume of training data, it is impossible to prepare these systems for every conceivable scenario. Thus, the potential for such errors always exists.

Finally, the third source of errors made by artificial systems is that machines lack morality and ethical point of views. In many fields, AI is being used to make decisions with moral implications, often without human oversight. For example, AI systems can identify and track individuals, potentially leading to biased or unjust actions. In some countries, AI algorithms are already being used in the justice system to assess the likelihood of a defendant committing future crimes, influencing decisions on bail, sentencing, and parole. These algorithms analyse various factors, from a person's criminal history to their socioeconomic background, to predict the risk level they pose. Without human supervision, AI can often misread the nuances of criminal cases and generate wrong verdicts. Even with human oversight, AI's decisions can influence a human judge's decision-making process in a wrong way.

An examination of Bangladesh's AI policy shows that, even though it addresses several potential sources of errors, some parts of it require further fine-tuning and expansion. For example, while the policy emphasises the need to prevent prejudice, bias, and discrimination in AI (Section 6.1.4), it lacks specific guidelines on ensuring diversity in training datasets, which is crucial for such prevention. Even though the legal and regulatory framework is meant to be established separately in the National Strategy for AI, which is a companion to the National AI Policy, more explicit instructions on data diversity, algorithmic oversight, and ethical considerations would benefit it.

The issue of data diversity is probably the most crucial for Bangladesh. Currently, most AI systems are being trained using datasets from developed countries. This can have profound negative implications for developing countries like ours. For example, in the near future, when AI becomes more commonplace, state-of-the-art AI algorithms used in medical diagnostics may not work well for patients in our country because the training dataset did not include enough information about our weather, food habits, and genetic makeup, which are pertinent to our health. Therefore, while data privacy and security are crucial, developing a representative demographic dataset under government supervision is also essential.

Bangladesh must ensure its AI system is fair, secure, and beneficial for all. While the draft National AI Policy marks a first step in that direction, policy gaps in data diversity, algorithmic oversight, and ethical considerations must be addressed more comprehensively in future revisions.

Amio Galib Chowdhury is graduate research student at the McCoy College of Business, Texas State University, US.​

 

[Saif]

DeepSeek’s AI triumph: A call for Bangladesh to retain its brightest minds

The global AI race is no longer just about talent migration—it is about fostering innovation at home.

www.thedailystar.net

DeepSeek’s AI triumph: A call for Bangladesh to retain its brightest minds

The recent success of DeepSeek, an AI model developed in China, has disrupted the industry, proving that world-class AI can be created at a fraction of the cost. FILE PHOTO: REUTERS

The United States has long been the world's premier destination for innovation, drawing the brightest minds from across the globe. With its unmatched financial resources, world-class research institutions, and a risk-taking culture that fuels technological breakthroughs, the US has cultivated an environment where ambitious thinkers thrive. The country has built an ecosystem that rewards talent and creativity, making it a magnet for scientists, engineers, and entrepreneurs who push the boundaries of possibility. This ability to attract and invest in global talent has been a key driver of the US's dominance in artificial intelligence (AI) and other cutting-edge technologies.

However, the landscape is evolving. The recent success of DeepSeek, an AI model developed in China, has disrupted the industry, proving that world-class AI can be created at a fraction of the cost. Unlike models from OpenAI or Google that require immense computational power and financial backing, DeepSeek demonstrated remarkable efficiency by utilising only a fraction of the GPUs typically needed for large-scale language models. This milestone underscores a broader shift: Asian countries are no longer just suppliers of talent for US companies; they are becoming AI powerhouses themselves. This shift is further reflected in the increasing presence of Asian talent in the US tech industry. With 71 percent of AI researchers in the US being foreign-born and a majority hailing from Asia, the statistics speak for themselves—35.9 percent of Microsoft's AI team and 44.8 percent of Google's AI researchers are of Asian descent.

What makes Asian talent so exceptional? The answer lies in the unique motivations that drive these individuals. Unlike their Western counterparts, who often have access to ample resources, many Asian innovators grow up in environments where they must navigate scarcity and intense competition. This fosters a problem-solving mindset, resilience, and an ability to maximise efficiency with limited means. In countries like Bangladesh, students develop an extraordinary work ethic from a young age, competing for a limited number of top university spots and thriving in challenging conditions. Despite the increase in higher education institutions, the quality of education remains a pressing issue. Many universities focus on infrastructure rather than investing in skilled faculty or research facilities, and there is no robust mechanism to measure educational quality. As a result, students seek better opportunities abroad, where they can specialise and gain practical knowledge tailored to industry demands.

Many of Bangladesh's brightest students have showcased their talents on the global stage, excelling in international competitions and pursuing higher studies at prestigious institutions abroad. A notable example of local ingenuity is Avro Keyboard, a free and open-source Bangla keyboard developed in 2003 by young innovator Mehdi Hasan Khan and his team. Over the years, it has become an indispensable tool for Bangla typing and was honoured with the 2025 Ekushey Padak in recognition. This innovation underscores Bangladesh's potential to contribute meaningfully to the technological landscape, demonstrating how the country's youth can drive innovation and make significant strides in the global tech arena.

Bangladesh, a country with immense intellectual potential, finds itself at a crossroads in this shifting global dynamic. The number of students going abroad for higher studies has tripled in 15 years, despite significant growth in local higher education institutions. In 2022, at least 49,151 Bangladeshi students went abroad to study in 58 countries, a sharp rise from 16,609 in 2008. Although the number of universities has doubled in the same period, the outflow of students has not slowed due to challenges such as limited research opportunities, outdated lab facilities, and political instability. The lack of skilled faculty, research funding, and infrastructure prevents them from thriving within the country, leading to a continuous brain drain.

To harness the potential of its bright population, Bangladesh must learn from countries that have successfully reversed the brain drain. China, for instance, implemented policies to bring back top researchers by offering competitive salaries, advanced research facilities, and government-backed funding programmes. Similarly, India's expanding AI and tech ecosystem is creating opportunities that encourage talent to return home. Bangladesh can adopt similar strategies by investing in AI research centres, fostering collaboration between academia and industry, and providing financial incentives for researchers to develop groundbreaking technology within the country. Creating a startup-friendly environment and strengthening AI infrastructure will allow Bangladesh to not only retain talent but also attract global investments.

The global AI race is no longer just about talent migration—it is about fostering innovation at home. While the US continues to benefit from Asia's intellectual capital, the balance is shifting. However, Bangladesh has a golden opportunity to leverage its human potential and become a significant player in AI. By investing in its brightest minds and providing them with the resources they need, the country can position itself at the forefront of the AI revolution. If Bangladesh embraces the momentum of innovation, it can transform from a talent exporter to a leader in global AI innovation.

Ruhina Tabasshum Prome is research associate at Bangladesh Institute of Governance and Management (BIGM).

Shirin Sultana is research associate at BIGM.​