PETALING JAYA — Safety and security assessments must precede the adoption and deployment of new technologies to prevent operational failures and safeguard user reliance, said Digital Minister Gobind Singh Deo.

While rapid technological advancements, ranging from robotics to Large Language Models (LLMs), offer significant opportunities, users and policymakers must understand the inherent risks, verification processes and control mechanisms required before full implementation.

Speaking at the MyMahir Artificial Intelligence (AI) Roadshow here today, Gobind highlighted a previous high-profile AI event launch that encountered technical glitches involving an autonomous robot, which provided crucial lessons on technological readiness and security controls.

“Actually, we must ensure that it (AI) is safe before we use it. So, we should be in a position where we are sure or certain that it is secure to use the technology before we introduce it and ask people to adopt it class-wide.

“Ultimately, there must also be layers to ensure that if anything goes wrong, we are still in a position to control it,” he added. 

Gobind noted that robust digital infrastructure and fail-safe layers are essential in robotics and autonomous deployment.

Addressing the surge in AI adoption, including tools such as ChatGPT and Gemini, Gobind cautioned users to verify the authenticity and accuracy of information before making critical decisions.