Why Bangladesh Must Prioritize AI Regulations Now

 

Artificial Intelligence (AI) has emerged as a defining factor shaping the future of governance, communication, and development. Its ability to mimic human intellectual functions, create content, and decide based on information is essentially transforming the working dynamics of societies. However, this rapid development is coming with an unprecedented degree of risk. The same technology that promises progress can also be used for intentions of deceit, manipulation, and destabilization. In Bangladesh, the misuse of AI has turned from a potential problem to a clear and present danger. As Bangladesh steps deeper into the digital age, it is imperative that we initiate serious endeavors in the governance of AI and take firm steps to regulate its use before it causes irreparable damage.

Over the past few months, a deluge of AI-generated content has inundated Bangladesh's social media. From fake news clips to doctored videos and cloned voices, the typical Facebook feed nowadays is not only a means of personal updates but also a war zone of misinformation. The circulation of fake political speeches, false news headlines, synthetic videos, and completely fabricated situations has started undermining public trust. Worryingly, a large number of people especially in rural and semi-urban settings cannot tell what is real and what is artificially generated. The outcome is confusion, fear, and at times unrest.

This is no speculation. The Chief Election Commissioner of Bangladesh has recently warned that AI could prove more dangerous than weapons in the electoral context. As elections are upcoming and online content is disseminating quicker than ever, the potential for AI-generated misinformation to sway public opinion or cause violence is quite real. In a time when Deepfake videos can falsely replicate political figures with convincing accuracy, it is becoming more challenging for the general public to believe what they view or listen to online. Such increasing lack of confidence can endanger not just our electoral process but also democracy's very foundations.

Across the world, alarm bells are also ringing. The UN's tech Chief has recently demanded a global response to AI regulation on an urgent basis. The United States and members of the European Union have already begun to act. The EU's Artificial Intelligence Act has brought in a risk-based approach to examine and regulate AI technologies. The United States has framed an AI Bill of Rights. India is readying its own AI governance guidelines. Bangladesh remains without a way forward.

At present, while Bangladesh lacks a fully developed framework for the regulation of artificial intelligence, important steps have already been taken in the right direction. Professor Asif Nazrul noted in a May press briefing that the recently passed Cyber Security Ordinance 2025 has achieved a noteworthy milestone by being the first piece of legislation in South Asia to specifically recognize crimes involving artificial intelligence as punishable offences. This is a crucial first step in recognizing the risks posed by AI to our legal system. However, given the increasing complexity of AI misuse, the ordinance is insufficient on its own. There are still significant gaps in domains like synthetic media, algorithmic discrimination, and deepfake regulation. Bangladesh has yet to establish an independent regulatory body dedicated to AI oversight, provide clear ethical guidelines for developers, or create public mechanisms for verifying AI-generated content.

The threat of AI is not just political. We are also witnessing its exploitation in economic scams, cybercrime, and even in fields such as education and healthcare. AI-created bogus job circulars and fraudulent scholarship notices have deceived many, especially youth and job-seekers. Cloned voices have been used over the phone to impersonate relatives or authorities. In schools and universities, students are resorting to AI-completed assignments and research papers, compromising academic integrity. Deepfake images and videos are being used with growing frequency to harass individuals, especially women, in ways our current laws are not prepared to tackle.

In cyber-security, AI is equipping hackers with more intelligent phishing campaigns and automated attacks that can evade traditional defences. However, as Bangladesh embraces more digital platforms for governance and finance, the vulnerability to AI-augmented cyber-attacks increases. AI also facilitates the production of counterfeit identities and bots that inundate comment sections or rig online trends, distorting public discourse and undermining democratic discussion. Here, it is apparent that Bangladesh cannot afford to view AI as solely a technological matter—it is a national governance and security dilemma.

The government's ongoing digital transformation including a digital economy, e-governance, and automation of education and healthcare, will not be successful unless it is founded on trust and integrity. That trust cannot flourish in a digital space filled with deceptive AI-created content. It is therefore essential to have a governance structure for AI in place urgently. That structure should not be intended to stifle innovation but should ensure that such innovation is secure, ethical, and beneficial to society as a whole.

The government needs to constitute a National AI Ethics and Safety Commission, which would have representatives from law, technology, education, media, and security. This commission could evaluate risks, supply guidelines for ethical use of AI, and assist in developing legislation. New sanctions need to be developed for the hostile use of AI-generated content, particularly in political and communal life.

Another key measure is requiring watermarking or explicit labelling of AI-generated content, particularly images, videos, and audio clips. Large global technology firms already have efforts underway to incorporate watermarking mechanisms. Bangladesh can enact the same requirements through local media channels and platforms. Large public awareness drives are also necessary to enhance AI literacy among all sections of society. Individuals need to be educated on how to identify fake videos, authenticate sources, and critically evaluate content online.

Education and bureaucracy must also evolve. AI literacy needs to be included in secondary and tertiary education curricula. Public officials, educators, and journalists must be trained on how to engage responsibly with AI tools and detect AI-based manipulation. The government, on its part, must foster ethical AI innovation through grants and cooperation with the private sector to nudge research into socially useful directions. It is indisputable that AI can be a forceful facilitator of national progress. If utilized judiciously, it can revolutionize education, transform agriculture, refine disaster management, and reimagine governance. However, if used unwisely or not regulated, it may undermine our democratic processes, intensify social fragmentation, and destabilize public confidence. Much is at stake.

Bangladesh is at a crossroads. We can either remain inactive in the face of technological disruption or take decisive and responsible action to create a safer, smarter tomorrow. With international leaders demanding action now and local threats already apparent, we cannot afford to wait. AI is not waiting. Neither should we.


Md. Abdullah Al Musa completed higher
studies in International Relations, 
Dhaka University.