Malaysia is proposing an AI Governance Bill that acts as a preventive layer across the entire life cycle of AI systems, addressing risks from deepfakes and synthetic content to broader safety.

The plan combines safety-by-design during development with stronger enforcement for content that breaks laws, aiming for a balanced approach to innovation and risk control.

Digital Minister Gobind Singh Deo notes that while safeguards exist for crimes against children, the legal framework must evolve to encompass AI-related risks.

The government will actively seek input from academics, religious leaders, and other stakeholders to inform the legal review process.

Enforcement will remain aligned with existing statutory boundaries, ensuring that new governance supports rather than hinders innovation.

The consultation process includes academics and religious leaders to ensure regulations stay effective as AI evolves.

The aim is to balance regulating emerging technologies with maintaining functional enforcement under current statutes.

Content that is illegal, such as child sexual exploitation material, would continue to be prosecuted under existing laws like the Sexual Offences Against Children Act 2017 and related Acts.

Existing laws will be enhanced to keep pace with technology, enabling investigation and prosecution of AI-related offences within the current legal framework.

The government acknowledges risks can arise before content is produced, including during design, data selection, training, security, oversight, and use.

Malaysia’s approach reflects a broader international challenge of adapting legal systems to AI harms, signaling a trend toward combining prevention, risk management, and enforcement.

The framework targets high-capability technologies like deepfakes and identity manipulation at an early stage with a layered, preventative approach.