Draft by Sept to enact law checking AI misuse: Anisul

To leverage artificial intelligence (AI) in economic development while ensuring its ethical applications, the government is working to formulate the draft AI Act by September, said Law, Justice and Parliamentary Affairs Minister Anisul Huq.

The law should be in line with the draft of National Artificial Intelligence Policy 2024 formulated earlier this year, he said at a stakeholders' discussion meeting on the upcoming AI law in the capital today (4 April).

Four priorities – regulation, innovation, facilitation, and collaboration – would be reflected in the law, said State Minister for ICT, Post, and Telecommunication Zunaid Ahmed Palak, who presided over the meeting.

In the era of the fourth industrial revolution (4iR), AI brings enormous opportunities for Bangladesh as well, while the risk of AI misuse is the main concern for all, said Palak, adding that the Act will protect citizens from unethical applications of AI as well as opening doors for proper regulations, encouraging innovations, facilitating AI utilisation and fostering collaboration.

"We can use AI for more accurate and efficient work. On the other hand, many crimes are taking place by using deepfake voice or video content that has security threats," he said.

For instance, a recent video of a newly built ICT park building, presenting it as an abandoned one later, was found to have been created by deep faking.

In addition to protecting users from harmful content, the AI Act would also help punish wrongdoers.

Palak also called for increased public awareness about the risks of AI abuse.

Anisul Huq stressed the need for drafting the Act in a way that overregulation does not limit the potential of leveraging AI, as the country is chasing the Smart Bangladesh dream by 2041.

Stakeholders called for reflection on the six international principles of AI regulations – human agency and oversight, technical robustness and safety, privacy and data governance, transparency, diversity, non-discrimination, and fairness, and social and environmental well-being.

There also needs to be harmony with AI laws in the major countries and regions like the US, the European Union, and others to avoid conflicting situations in AI practices in the future, said stakeholders.

The proposed National AI Policy 2024 focuses on AI adoption and leveraging it to build a Smart Bangladesh that upholds the wellbeing of all citizens, economic prosperity, and sustainable development.

Its six principles are social equity, equality and fairness, transparency and accountability, safety, security and robustness, sustainability, partnership and collaboration, and human-centered AI.

The policy said, AI should be integrated in priority areas – smart public services, governance and judiciary, telecommunication, data governance and surveillance, agriculture, environment, energy, and climate change; smart cities; transportation and mobility; smart finance; trade and economy; smart manufacturing and industrial transformation; education; skill development and employment; healthcare and wellbeing; and science, technology, research, and innovation.

The policy outlined the institutional framework for implementing the AI Act, challenges and mitigation, risk management, and, most importantly, the evolution of the policy itself.

Some 26 stakeholders, representing industry and academia, shared their thoughts at the meeting.

Bangladesh Telecommunication Regulatory Authority Chairman Md Mohiuddin Ahmed, ICT Secretary Md Shamsul Arefin, and Law Secretary Md Golam Sarwar also spoke on the occasion.