DOI:
AUTHOR(S)
Zulfiquar Ahmed
ABSTRACT
This article examines the regulatory challenges posed by artificial intelligence in Bangladesh, with a critical focus on the Cyber Protection Ordinance, 2025. Using a hybrid doctrinal method, the study traces digital law’s evolution from the ICT Act 2006 to CPO 2025, examining how the current framework addresses AI challenges such as accountability, automation, and rights protection within a rights-based constitutional framework. It identifies significant doctrinal gaps in the CPO especially in the absence of a functional or risk-based definition of AI, insufficient liability attribution mechanisms, and lack of enforceable ethical safeguards. The article also considers soft law instruments like the National AI Policy 2024 (Draft), AI Readiness Report 2025, and the National AI Strategy. It finds that although CPO 2025 marks progress, it lacks sufficient normative depth and institutional support for effective AI governance. By incorporating case studies of AI misuse (e.g., deepfakes, algorithmic bias in public services) and constitutional analysis (Articles 27, 31, and 39), the article situates Bangladesh’s challenges within both domestic and global contexts. The article contributes to AI legal learning by offering concrete reform proposals to ensure a human-centric, rights-based regulatory regime aligned with global standards.
Download Full Article