A Comparative Legal Analysis of Deepfake Evidence in Bangladesh

References

[1] 5KBW. (2025, June 30). Sexual offences in the Online Safety Act of 2023. https://www.5kbw.co.uk/2024/11/28/sexual- offences-in-the-online-safety-act-2023/

[2] Ahmad Shariff, N., Ismail, R., & Hamzah, N. (2021). Challenges of the admissibility of digital evidence in Malaysia. Current Law Journal, 9(2), 1–18.

[3] Chesney, R., & Citron, D. K. (2019). Deepfakes: A looming challenge for privacy, democracy, and national security. California Law Review, 107(6), 1753–1820.

[4] Council of Europe. (2022). Artificial intelligence and the criminal justice system. Council of Europe Publishing,

[5] The Evidence Act of 1872 (Act No of. I of 1872) (Bangladesh).

[6] Farid, H. (2022). Creating and detecting deepfakes. Springer.

[7] Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial networks. Advances in Neural Information Processing Systems, 27, 2672– 2680.

[8] Government of Bangladesh. (2022). Evidence Act, 1872 (Amendment of 2022). Ministry of Law, Justice and Parliamentary Affairs.

[9] Government of Bangladesh. (2025). Cybersecurity Act of 2026. Ministry of Law, Justice and Parliamentary Affairs.

[10] Hansard (UK Parliament, House of Lords). (2026, January 14). Social media: Non- consensual sexual deepfakes [Debate transcript]. https://hansard.parliament.uk/lords/2026-01- 14/debates/2AF08B6C-1AFC-4395-B356- E383ABF28161/SocialMediaNon- ConsensualSexualDeepfakes

[11] Lexplosion Solutions. (2025, December 22). Cyber Security Ordinance 2025: Implications for Bangladeshi Businesses https://lexplosion.in/cyber-security-ordinance- 2025-implications-for-businesses-in- bangladesh/

[12] Malaysia. (1950). Evidence Act of 1950 (Act No.56).

[13] Mason, S., & Seng, D. (Eds.). (2021). Electronic evidence and electronic signatures (6th ed.). Institute of Advanced Legal Studies, London, UK.

[14] Mohamad, D. (2019). Digital forensics and the admissibility of electronic evidence in Malaysian Syariah courts: Towards a standardized approach. Lex Forensica Journal, 1(1), 1–22.

[15] National Institute of Standards and Technology (NIST). (2023). Reducing the threat of deepfake manipulation: U.S. Department of Commerce,

[16] Prothom Alo. (2025, May 22). The government issues a gazette on the Cyber Security Ordinance. https://en.prothomalo.com/bangladesh/govern ment/5gj5obvmuw

[17] QMUL Legal Advice Centre. (2026, March 31). Deepfakes and consent: The law finally catches up. Queen Mary University of London, London, UK https://www.qmul.ac.uk/lac/our-legal- blog/2026-blogs/items/deepfakes-and- consent-the-law-finally-catches-up-.html

[18] Rahman, M. T., & Naser, A. (2023). Admissibility of electronic evidence in Bangladesh: Contemporary challenges and prospects. Bangladesh Journal of Law, 34(1), 45–63.

[19] Roberts, P., & Zuckerman, A. (2022). Criminal evidence (3rd ed.). Oxford University Press.

[20] Russell, S., & Norvig, P. (2021). Artificial intelligence: A modern approach (4th ed.). Pearson.

[21] TBS News. (2026, June 8). Cyber Security Act to be amended to remove harmful AI- generated content: Home Minister. https://www.tbsnews.net/bangladesh/cyber- security-act-be-amended-remove-harmful-ai- generated-content-home-minister-1457531

[22] UNESCO. (2023). Guidance for the governance of digital platforms. UNESCO Publishing.

[23] United Kingdom Government (UKG). (1984). Police and Criminal Evidence Act, 1984.

[24] United Kingdom Government (UKG). (2020). Criminal Procedure Rules 2020.

[25] United Kingdom Government (UKG). (2023). Online Safety Act 2023.

[26] United Kingdom Government (UKG). (2025). Data (Use and Access) Act of 2025.

@article{1722738,
      author = {MD Shamsad Hossain},
      title = {A Comparative Legal Analysis of Deepfake Evidence in Bangladesh},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {10},
      number = {3},
      pages = {249-262},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1722738.pdf},
      abstract = {Deepfake technology, which involves the creation of synthetic audio, video, and image content through artificial intelligence to portray real individuals engaging in actions or speech that never occurred, poses a significant evidentiary challenge that the traditional digital-evidence doctrine is not equipped to handle. Bangladeshi courts are increasingly dependent on electronic records; however, the Evidence Act of 1872, even after its amendment in 2022, lacks provisions addressing AI-generated content, authentication standards specific to synthetic media, and qualification frameworks for experts required to identify such content. This study explores whether Bangladesh's evidentiary framework is sufficient to meet this challenge and, if not, what reforms are necessary to bridge this gap. Utilizing doctrinal and comparative methods, this paper analyzes the statutory and case law treatment of digital evidence in Bangladesh in comparison with the approaches adopted in the United Kingdom and Malaysia. These jurisdictions were chosen for their shared evidentiary heritage and differing regulatory styles: judicial reliability review in the UK and codified statutory presumption in Malaysia. The analysis reveals that none of the jurisdictions reviewed have established a deepfake-specific evidentiary regime, but each provides transferable doctrinal elements: Malaysia's structured certification mechanism, the UK's substantive reliability inquiry, and an emerging legislative trend, as seen in the Cyber Security Act of 2026 in Bangladesh, which considers the direct regulation of AI-generated content. This study proposes a sequential authentication framework and a series of legislative, institutional, and procedural reforms designed to be integrated within Bangladesh's current judicial infrastructure rather than relying on it alone.},
      keywords = {Deepfake evidence, artificial intelligence, digital evidence authentication, Evidence Act 1872, admissibility, comparative evidence law, digital forensics.},
      month = {September},
      doi = {https://doi.org/10.64388/IREV10I3-1722738}
  }