Govt says AI chatbots aimed at children will have safety built in

Developers of AI chatbots and apps targeting children will be required to incorporate safety features under the proposed Artificial Intelligence (AI) Governance Bill, says digital minister Gobind Singh Deo.

In a written parliamentary reply, Gobind said child protection was being “comprehensively” integrated into the bill, which was being drafted as a foundational framework that could adapt to emerging technology risks.

He said the bill would take a risk-based approach, requiring developers of apps or chatbots aimed at children to incorporate “safety-by-design” measures.

“This includes requiring regular risk impact assessments and filtering harmful content to prevent the psychological manipulation of children,” he said.

Gobind said bot disclosure – notifying users that they are interacting with an AI system rather than with a human – would be enforced through a “targeted mechanism” rather than a blanket requirement.

He said the need for disclosure would be determined based on an application’s risk level, sector-specific requirements, and specific laws or sectoral codes of ethics to be issued by the regulator proposed under the bill and by the minister.

Under this phased approach, AI service providers would eventually have to display clear notices informing users that they are interacting with AI, particularly for applications classified as high-risk, he said.

Gobind said the bill’s legal controls would be reinforced through the National Guidelines on AI Governance and Ethics, enforcement of the Online Safety Act, and digital literacy initiatives such as “CyberSAFE untuk Rakyat”, a free cybersecurity awareness programme.

He was responding to a question from Riduan Rubin (independent-Tenom) on the safeguards in the AI Governance Bill to mitigate the risks of children using AI chatbots and whether the government would make bot disclosure mandatory to protect minors’ cognitive awareness.

Yesterday, Gobind said the government expected to table the AI Governance Bill in the Dewan Rakyat early next year instead of this year.

He said the government was still drafting the bill, and the level of control and obligations imposed would be assessed based on the potential harm posed by an AI system.