Artificial Intelligence (AI) has rapidly changed digital life through its applications in content creation and automated processes and communication systems. The technology has created a new type of criminal activity which law enforcement agencies now call AI-generated crimes. The tools which include deepfake videos and voice cloning together with automated scams can be used to create false information for the purpose of damaging others and committing fraud and causing distress to victims. In Bangladesh, the legal response to such conduct is primarily grounded in the Cyber Security Ordinance, 2025 that is basically supplemented by traditional criminal laws.
The Cyber Security Ordinance 2025, Sections 21 and 22 define impersonation and identity theft crimes which AI technologies help criminals to commit. The regulations apply to social media accounts that users create for deception purposes together with users who alter images and videos and engage in online scams. The actual implementation of these systems includes deepfake technology for identity theft and artificial identity creation fraud schemes. The law gives imprisonment sentences that extend to five years for offenders who commit serious crimes while imposing fines that can reach BDT 50 lakh. The Ordinance addresses wider societal damages beyond its primary focus. Section 26 targets hate speech because it includes AI-generated material which incites religious, ethnic and communal violence. Section 27 expands legal responsibility to individuals who help others commit crimes because it includes all forms of organized trolling and stalking activities. Section 29 establishes platform responsibility through which intermediaries must prove their systems contained sufficient security measures before they can avoid liability for offenses that happened on their platforms. Section 30 enables courts to grant compensation for mental distress and reputational harm and financial damages in addition to criminal penalties.
Digital environments still operate under established traditional criminal law. The Penal Code 1860 defines defamation in Section 499 through any imputation which defenders intended to destroy a person's reputation, including all AI-created deepfakes and falsehoods. Section 500 prescribes punishment of up to two years' imprisonment, or fine, or both. The sections 503 and 506 laws deal with criminal intimidation through threats which people make on online platforms. The laws establish that digital media should maintain legal frameworks.
“The Cyber Security Ordinance 2025, Sections 21 and 22 define
impersonation and identity theft crimes which AI technologies help
criminals to commit. The regulations apply to social media accounts that
users create for deception purposes together with users who alter
images and videos and engage in online scams. The actual implementation
of these systems includes deepfake technology for identity theft and
artificial identity creation fraud schemes”
The Pornography Control Act, 2012 establishes particular regulations which apply to situations where artificial intelligence technology creates or disseminates explicit material without obtaining user approval. The law prohibits all activities related to the creation and distribution and storage of obscene content which serves as a vital instrument for combating deepfake pornography and other types of sexual exploitation which primarily impact women and other vulnerable populations.
The existing legal framework fails to provide effective enforcement solutions. The main issue with AI-generated material is its ability to create highly authentic content which users find difficult to trace back to its source. The presence of anonymous offenders who commit crimes while using cross-border operations makes it difficult to determine their identity and the applicable legal jurisdiction. The Bangladesh Telecommunication Regulatory Commission together with specialized cybercrime units performs crucial functions yet their ability to investigate and prosecute cases suffers from digital forensics capacity limitations and their lack of technical skills.
Normative issues also exist. Some statutory provisions have an expansive meaning and can be interpreted differently, making careful scrutiny by courts necessary to avoid any abuse. This point becomes even more relevant when one considers the provision provided in Article 39 of the Constitution of Bangladesh regarding the freedom of expression and thought. There is a need for a delicate balance here. Public awareness represents another a gap.
Firstly, it is recommended to provide specific legal definitions of AI, deepfakes, and AI-generated materials in the Cyber Security Ordinance, 2025 to avoid misinterpretation. Secondly, it is necessary to strengthen the forensic capacity through education and procurement of AI detection software. Thirdly, it is proposed to enhance the responsibility of platforms and create transparency regarding takedown actions. Fourthly, efforts should be made to increase public awareness and assist victims and paying special attention to the support of vulnerable categories like women. Fifthly, it is required to develop international cooperation in the case of transnational crime.
The increasing instances of AI-driven crime in Bangladesh are becoming a serious concern for law enforcement agencies. The provisions provided under the Cyber Security Ordinance, 2025, together with those provided under the Penal Code, 1860, and the Pornography Control Act, 2012, provide a framework through which the issue can be legally addressed. However, the issue is complicated by factors such as ambiguity in definitions, low forensic capabilities, cross-border anonymity, and concerns about freedom of expression. Hence, it calls for more stringent legislation, improved cybercrime investigation technology, platform accountability, public enlightenment, and international collaboration.
The writer is a student, BUP
