The rapid rise of artificial intelligence (AI) has revolutionized various aspects of daily life, from automated manufacturing to generative AI applications like ChatGPT. However, the portrayal of AI in movies such as The Terminator and TV shows such as Westworld, has fueled concerns about its potential negative impact. While the dystopian scenarios depicted in fiction are unlikely to occur in the near future, AI, if kept unsupervised by a wider regulatory framework, poses real-world risks, including the spread of misinformation, job displacement, and manipulation of individuals’ lives1. Given the swift evolution of AI technologies, there is a pressing need to discuss and implement regulatory frameworks to mitigate potential adverse effects.
The CEO of OpenAI, Sam Altman, warned US Congress that AI can go “quite wrong” and has emphasized on the urgent need for government to create regulations surrounding AI.2 After its launch, OpenAI’s ChatGPT became the fastest growing website in history, garnering over 100 million users in two months. This success triggered a competitive surge, with companies like Microsoft, Google, Meta, and IBM investing billions in developing their own AI models. As numerous companies and developers globally join this AI race, it is important for governments to establish and develop strict rules, and enforce legal guardrails to prevent reckless competition and ensure responsible AI development.3
Prospects of AI in Bangladesh
“Digital Bangladesh”, a slogan by the Government of Bangladesh, is of special significance for national development, and works have been in progress to increase digitalization of the country to realize this initiative.
With the government expressing its commitment to embrace technology and innovation, the prospects of AI in Bangladesh are promising. There is increasing interest in adopting artificial intelligence across various sectors, and a growing awareness of the potential benefits of AI in areas such as healthcare, agriculture, finance, and education. As the potentials of AI are coming to limelight in different sectors, initiatives are being taken to integrate AI technologies to enhance efficiency, productivity, and decision-making processes.
The current progress in regulating AI – The Draft National Strategy for Artificial Intelligence of Bangladesh
So far, Bangladesh has introduced a draft policy paper titled “National Strategy for Artificial Intelligence of Bangladesh”. The country has recognized the potential of AI to drive digitalization and improve productivity across various sectors, including public service delivery, manufacturing, agriculture, smart mobility and transportation, skill & education, finance & trade, and health. To ensure the responsible and ethical implementation of AI, the National Strategy for Artificial Intelligence of Bangladesh has identified the need for legal regulations to establish an ethical and legal framework for AI. The strategy emphasizes the importance of formulating data sharing policies, making laws for data ethics and privacy, ensuring AI security, and developing a Right to Explanation (RTE) guideline for AI algorithms. The strategy also highlights the need for sensitizing policymakers, institutionalizing the legal framework, and enforcing laws and policies to promote awareness and ensure effective implementation of these legal regulations for AI in Bangladesh.4
The National Strategy for Artificial Intelligence further recognizes that the adoption of AI poses potential challenges and risks, such as the shortage of skilled AI resources, economic impacts, accountability, transparency, privacy, and legal and ethical frameworks. To address these challenges, the strategy aims to promote capacity building, policy development, and stakeholder engagement. It includes a detailed roadmap for the development and implementation of AI in Bangladesh, with specific action plans, stakeholders, and lead ministries identified for each sector and pillar. It also emphasizes the importance of monitoring and regulation by various government ministries, including the Ministry of Law, Justice and Parliamentary Affairs, Cabinet Division, Ministry of Planning, and the ICT Division.5
How is AI being regulated in other parts of the world?
In the interest of determining the correct approach to regulate AI in Bangladesh, it can be helpful to study the existing and emerging legal frameworks in the international community, and the schools of thoughts they bring into regulating this double-edged sword that is Artificial Intelligence.
EU AI Act: First regulation on artificial intelligence
On 9 December 2023, European Parliament negotiators and the Council presidency agreed on the final version of what is claimed to be the world’s first-ever comprehensive legal framework on Artificial Intelligence – the European Union Artificial Intelligence (the “EU AI Act”). 6
The EU AI Act is a comprehensive legal framework with the aim to regulate the development, deployment, and use of AI systems in the European Union based on their level of risk to human health, safety, and fundamental rights.7 The EU AI Act also seeks to foster innovation and competitiveness in the AI sector, while ensuring that AI systems respect EU values and rules.8 To achieve these goals, the AI Act adopts a risk-based approach, categorizing AI systems into four groups: Prohibited, High-risk, Limited-risk, and Minimal-risk.
Prohibited AI systems, which compromise human dignity through actions like manipulating behavior or exploiting vulnerabilities, are strictly forbidden. High-risk AI systems, involving significant risks to health, safety, or fundamental rights, such as those used in biometric identification or healthcare, must adhere to stringent rules regarding data quality, transparency, human oversight, accuracy, robustness, and security. Limited-risk AI systems, posing some risk to users or consumers, like those generating or manipulating content, are required to provide clear information about their nature and purpose and allow users to opt out of using them. Minimal-risk AI systems, carrying little to no risk, are encouraged to follow voluntary codes of conduct and best practices, particularly in entertainment or personal applications.
The European Union AI Act also aims to establish a governance structure for the implementation and enforcement of its rules. This includes a European AI Board (EAIB) that will provide guidance and advice on various aspects of the European Union AI Act, such as harmonized standards, codes of conduct and risk assessment methods. The EAIB will also facilitate cooperation and coordination among national competent authorities who will be responsible for monitoring and supervising compliance with the European Union AI Act in their respective territories.9
Canada and the United Kingdom
Canada, through its proposed Artificial Intelligence and Data Act (AIDA), takes a similar approach to the European Union.10 Canada will not ban any AI applications outright and will instead require AI developers to establish mechanisms that minimize risks and improve transparency, ensuring AI applications respect anti-discrimination laws and that their decision-making processes are clear.11
On the other hand, the UK Government’s AI White Paper, released on March 29, 2023, proposes a distinct regulatory approach for artificial intelligence (AI) compared to the EU’s AI Act. Instead of new comprehensive legislation, the UK aims to set expectations and empower existing regulators like the Information Commissioner’s Office (ICO), the Financial Conduct Authority (FCA), and Competition and Markets Authority (CMA) to guide and regulate AI use.12 The White Paper defines AI based on adaptivity and autonomy and outlines three pillars: investing in the AI ecosystem, ensuring benefits across sectors, and effective governance. The strategy also outlines specific actions to achieve these pillars, such as launching a draft National Strategy for AI in Health and Social Care, building an open repository of AI challenges with real-world applications, and publishing research into the determinants impacting the diffusion of AI across the economy.13
China
China has enacted many AI relevant regulations since 2021, including a law for personal data protection, an ethical code for AI and most recently guidelines on the use of generative AI.14 Chinese laws grant users transparency rights to ensure they know when they interact with AI-generated content and the option to switch off AI recommendation services. Measures against ‘deepfakes’ – AI-generated content that is realistic but false – are also in place. However, many of the existing laws only apply to private companies that use AI and not to the Chinese state.15
Prospective regulations that can be introduced in Bangladesh
As of now, there is no specific law that regulates artificial intelligence (AI) in Bangladesh. Existing laws and regulations apply to AI technologies to limited extents. For example, the Digital Security Act 2018 primarily focuses on regulating digital devices and networks, addressing issues related to cybersecurity, data protection, and digital offenses.16 While the Act plays a crucial role in governing aspects of digital technology, it lacks specific provisions directly addressing the regulation of artificial intelligence (AI) technologies.
As Bangladesh strides towards harnessing the promising prospects of AI, it stands at a crucial juncture to shape responsible regulations for its development and application. Building on the foundation of the Draft National Strategy for Artificial Intelligence, Bangladesh can draw inspiration from global regulatory frameworks to introduce prospective regulations tailored to its specific needs.
For instance, Bangladesh’s Draft National Strategy for Artificial Intelligence focuses on a general approach to outline the country’s priorities and goals for AI development, whereas the EU AI Act will provide a more detailed and specific regulatory framework for AI technologies. A key consideration could be adopting a risk-based approach, categorizing AI applications based on their potential impact and implementing corresponding rules for transparency, accountability, and oversight. The establishment of a dedicated regulatory body such as the European AI Board (EAIB), as seen in the EU AI Act, could facilitate guidance, standardization, and enforcement of AI regulations. Introducing clear data sharing policies, robust laws for data ethics and privacy, and a Right to Explanation (RTE) guideline for AI algorithms, as outlined in the National Strategy, would further ensure ethical AI practices. Collaborative efforts among ministries, emphasized in the strategy, should be strengthened to enable effective governance and enforcement. Additionally, investing in capacity building, fostering stakeholder engagement, and regularly updating regulations to keep pace with technological advancements are crucial components for a comprehensive and adaptive regulatory framework. By studying the existing problems , and by assessing future risks, Bangladesh can integrate more robust regulations alongside it’s National Strategy for Artificial Intelligence and position itself as a responsible and forward-thinking participant in the global AI landscape.
Written by Junayed Ahmed Chowdhury (Partner) and Raed Saadman Karim (Pupil) at Vertex Chambers
† Disclaimer: The opinions and comments expressed in this Blawg are not to be regarded or construed as legal advice by and from Vertex Chambers or any of its members. It is highly advisable that any person should seek independent legal advice before relying on any of the contents of this Blawg.
[1] https://www.independent.co.uk/tech/congress-ai-chatgpt-sam-altman-b2340147.html
[2] ibid
[3] https://www.brookings.edu/articles/the-three-challenges-of-ai-regulation/
[4] Draft Version 3:0 National Strategy for Artificial Intelligence Bangladesh,2019-2024; https://ictd.portal.gov.bd/sites/default/files/files/ictd.portal.gov.bd/page/6c9773a2_7556_4395_bbec_f132b9d819f0/Draft%20-%20Mastering%20National%20Strategy%20for%20Artificial%20Intellgence%20-%20Bangladesh.pdf
[5] https://www.europarl.europa.eu/news/en/press-room/20231206IPR15699/artificial-intelligence-act-deal-on-comprehensive-rules-for-trustworthy-ai
[6] Edwards, L.; The EU AI Act: A Summary of Its Significance and Scope, Ada Lovelace Institute, UK, April 2022
[8] https://www.isaca.org/resources/news-and-trends/industry-news/2023/understanding-the-eu-ai-act#6
[9] The Artificial Intelligence and Data Act (AIDA) – Companion Document’ (Innovation, Science and Economic Development Canada 2023) <https://ised-isde.canada.ca/site/innovation-better-canada/en/artificial-intelligence-and-data-act-aida-companion-document
[10] https://www.adalovelaceinstitute.org/report/regulating-ai-in-the-uk/#_ftn33
[11] https://www.mayerbrown.com/en/perspectives-events/publications/2023/07/uks-approach-to-regulating-the-use-of-artificial-intelligence
[12] https://www.gov.uk/government/publications/national-ai-strategy
[13] Translation: Measures for the Management of Generative Artificial Intelligence Services (Draft for Comment) – April 2023’ (DigiChina) <https://digichina.stanford.edu/work/translation-measures-for-the-management-of-generative-artificial-intelligence-services-draft-for-comment-april-2023/>
[14] https://www.adalovelaceinstitute.org/report/regulating-ai-in-the-uk/#_ftn33
[15] ডিজিটাল নিরাপত্তা আইন, ২০১৮, <http://bdlaws.minlaw.gov.bd/act-1261.html?lang=bn>
AstuteHorse