As the use of Artificial Intelligence (AI) technology is becoming more commonplace, experts have lauded the government’s move to regulate it.
However, they also cautioned that it should not be done hastily as sufficient research was required for the purpose.
Senior Civil Law lecturer at the Ahmad Ibrahim Kulliyyah of Laws, International Islamic University of Malaysia (IIUM), Dr Mahyuddin Daud said the rapid development of AI could be seen as a double-edged sword, where it could help propel numerous industries in Malaysia such as healthcare, security, education and communication, but could also prove detrimental if there was no legal framework to oversee its autonomy.
Thus, he said establishing a framework for some of the concerns regarding AI, including the risk it posed to data privacy, unethical decision making and lack of accountability should be tabled.
"The use of AI has raised various ethical and moral questions and dilemmas. For example, the use of AI in weaponry and surveillance systems can pose significant risks to human rights and global security, if not properly governed by the law,” he told Bernama.
He said AI algorithms could also inadvertently create situations of bias and discrimination if not properly designed and monitored.
“For example, in a court that uses AI to determine punishment, if the algorithm is set not based on the principles of law and justice, AI may set the same sentence for two cases with different merits," he said.
Mahyuddin said considering the complexity and autonomous characteristic of AI, there was also a warranted urgency to have it regulated to ensure users’ interests were safeguarded by minimising the risk of unfortunate incidents occurring from the usage of ill-intended or badly designed code.
“In the context of AI-driven products and services, a legal framework can protect consumers from misleading or harmful practices. It can set standards for transparency in AI-generated content and require clear disclosure (transparency) when interacting with AI systems,” he said, adding that the legal framework will ensure clear accountability and liability when the use of AI systems causes harm or error.
Meanwhile, Universiti Malaysia Sabah’s Creative Advanced Machine Intelligence Research Centre director Assoc Prof Dr Rayner Alfred said there were several main areas of concern regarding AI including the lack of transparency of AI tools in which the decisions were not always intelligible to humans.
He said AI-based decisions were susceptible to inaccuracies, discriminatory outcomes, or embedded bias, whereby these uncertainties could affect human safety, privacy, and ethical governance.
As such, he said regulations regarding AI could be implemented by exploring the impact of outcomes generated by AI, the nature and scope of the works involved as well as taking into account its operational complexity and its self-compliance and governance capabilities.
“Any possible outcomes that can affect safety, fairness, and privacy should be mitigated earlier. In addition to that, if the operational complexity is too high, then someone or the AI programme should be able to explain the possible outcomes of the AI algorithms.
Rayner, who is also a board member of MIMOS Bhd, proposed that data privacy and protection regulations should also be strengthened before using AI algorithms, which were highly dependent on training or learning from users’ data.
“The development of an ethical AI framework should be proposed as there is no comprehensive Federal legislation dedicated to AI regulation now, and one should address the copyright and intellectual property concerns as AI could be reused for other applications,” he added.
On Tuesday, Communications and Digital Minister Fahmi Fadzil said the government was looking into the need to establish a regulatory framework for AI in order to understand some of the challenges of using the new technology, as well as to address ethical issues related to its use.
Source: BERNAMA News Agency
