Barriers to embracing AI

In Bangladesh, numerous negative stories exist aimed at discrediting AI and discouraging its adoption. One school introduced AI to grade Bangla essays. A student wrote, "My cow is my best friend." The AI flagged it as an inappropriate relationship and gave zero. The student cried, the teacher protested, and the AI resigned via email: "Cows confuse me emotionally." Meanwhile, a matchmaking agency used AI to pair brides and grooms. It matched a 22-year-old MBA graduate with a 65-year-old retired politician, citing "shared leadership qualities and love for microphones." The AI is now reportedly applying for jobs abroad.

According to a 2023 Gartner survey, 79 percent of corporate strategists stated that the use of AI, automation, and analytics would be crucial to their success over the next two years. But only 20 percent of them reported using AI in their daily activities. In the Bangladesh context, these numbers would be much lower.

Despite the global buzz, AI remains more of a boardroom aspiration than an operational reality in Bangladesh. While policymakers speak of "Smart Bangladesh" and the private sector toys with AI-led marketing or customer service, actual implementation faces resistance. Not just technical or financial resistance, but human resistance rooted in perception, culture, and fear. To unlock AI's potential, we must first address why people, whether garment workers or C-level executives, are hesitant.

Julian De Freitas, assistant professor at Harvard Business School, explores this very dilemma in his Harvard Business Review article, "Why People Resist Embracing AI." While his analysis focuses on psychological barriers, these insights apply strikingly well to Bangladesh, with some added layers of local complexity. Drawing from his work and extending it with regional realities, we can identify four major roadblocks.

First, there is the issue of trust and transparency. AI tools often produce results without explaining the rationale. In Bangladesh, where institutions already suffer from a trust deficit, opaque algorithms worsen suspicion. Imagine a farmer denied a digital loan or a student scored by AI with no explanation, confidence quickly erodes. The study suggests comparative explanations can increase trust. For Bangladesh, simplifying the logic behind AI decisions in Bangla and offering side-by-side reasoning is essential.

Second, AI is often perceived as culturally detached. Many AI tools are trained on Western datasets. As a result, they frequently misinterpret local dialects, customs, or user behaviour. A rural user from Jashore might abandon an AI-powered agricultural assistant that fails to understand her spoken queries. Localising data, training models on Bengali and regional variations, and incorporating cultural norms into design are non-negotiables for widespread adoption.

Third, the fear of job loss remains a dominant concern. In an economy reliant on labour-intensive sectors like RMG, logistics, and retail, automation can be viewed as a direct threat to livelihoods. But as De Freitas notes, this fear is often amplified when AI is portrayed as entirely replacing humans rather than supporting them. Bangladesh must reframe AI as a co-worker, helping RMG workers improve quality, aiding teachers in customising lessons, or assisting doctors with faster diagnostics. Reskilling initiatives and transparent deployment strategies will be crucial to calm anxieties.

Lastly, there's a significant emotional disconnect. Bangladeshis value personal warmth and familiarity in their day-to-day interactions. Emotionless chatbots or robotic interfaces can feel alienating. De Freitas argues that people are more open to AI when it mirrors human reasoning and empathy. Giving AI tools familiar names, incorporating polite and relatable language, and allowing seamless handovers to human support where needed can make all the difference.

AI in Bangladesh doesn't face a technological problem; it faces a human one. By understanding these psychological and cultural frictions and designing systems that inform, include, and empower users, we can turn cautious curiosity into confident adoption. AI, when made familiar, fair, and friendly, has the potential to become a true ally in our national journey toward a brighter, more inclusive future.

The writer is the president of the Institute of Cost and Management Accountants of Bangladesh and founder of BuildCon Consultancies Ltd.