PUTRAJAYA: A code of ethics on the use and development of artificial intelligence (AI) should not just be confined to government entities but also expanded to the private sector, says an expert.
The Artificial Intelligence Governance and Ethics Guidelines (AIGE), based on seven main principles, were launched here yesterday as a regulatory measure for responsible development and adoption of AI technology.
Dr Azree Nazri from Universiti Putra Malaysia’s Institute of Mathematical Research said the policy is pivotal in shaping the future of the country’s AI industry.
“With China’s overwhelming lead in generative AI patents and the complexities of effective regulation, Malaysia faces a challenging path.
“The AIGE offers a strong framework for responsible AI, but limiting it to government entities reduces its overall impact – as most AI development is led by private companies,” he said when contacted.
Azree – who is also Artificial Intelligence Society Malaysia president – said that without alignment from the private sector, AI practices could remain fragmented and undermine the goal of ethical governance.
“Incentives can help bridge this gap. Benefits such as tax breaks, grants and recognition programmes encourage companies to voluntarily adopt AIGE standards.
“Through voluntary adoption, collaboration and government incentives, AIGE’s goals of fostering fair, transparent and accountable AI can be achieved across both public and private sectors,” he said.
However, he said Malaysia had yet to have any local AI infrastructure.
“The AIGE is akin to establishing an airspace law but ‘without airports and airplanes’,” Nazree said.
He also said besides China, Malaysia needed to learn from the United States with regard to regulations governing the AI industry.
“In the US, regulators are grappling with how to manage the risks associated with high-capability AI systems. One approach involves setting thresholds for AI models based on their computational power.
“The move aims at addressing potential dangers without stifling technological progress, a strategy that Malaysia may need to consider as it develops its own regulatory framework,” Azree said.
Meanwhile, Cybersecurity consultant Fong Choong Fook said the AIGE was the best set of regulations the government could come up with to govern AI.
He said imposing ethics into law for AI would be very difficult for now due to the fluidity and variables of AI.
“With Malaysia just beginning to dabble in AI and with the future being bright for AI development and usage, this is a good time to set down such a code of ethics.
“This would enable the ethics to be ingrained from the start,” he said.
The seven AIGE principles are fairness; reliability, safety and control; privacy and security; inclusiveness; transparency; accountability and the pursuit of human benefit and happiness.
Under the fairness principles, the development of AI must be designed to avoid bias or discrimination against the target users.
According to the second principle, AI systems should perform as intended with consent for usage of data the top priority.
The principle of privacy and security also encourages constant, consistent testing and certification as well as risk and impact assessments.
All AI systems should not have one-size-fits-all approaches and should address the widest possible range of factors – such as age, gender, religion and ethnicity.
Transparency is pertinent in contexts where AI is used as part of decision-making and AI algorithms should be transparent – be they technical processes or human decisions.
Besides accountability, the other principle caters to respecting human-centred values and enhancing quality of life while ensuring that AI systems do not undermine human autonomy through governance mechanisms.
