Background

Malaysia’s recognition of the need for an ethical approach to artificial intelligence (AI) is embedded within the country’s National AI Roadmap (AI-Rmap). The AI-Rmap, drafted from December 2020 to March 2021, outlines the establishment of an AI Code of Ethics as one of four strategic initiatives that would contribute toward a broader AI governance framework.

The development of an AI Code of Ethics is projected to take place over a period of four years (2021–2025), with progressive yardsticks and measurable key performance indicators (KPIs) along the way.

Establishing an AI Code of Ethics

Horizon 1 (2021-2022)

Monitoring and analyzing ethical initiatives and impacts:

Benchmarking activities to develop a Code
Observing trending codes in international organizations and major countries
Drafting an AI Code of Ethics and Guidelines
Formulating AI ethical standards consistent with evolving global norms
Setting up a Center for Data Ethics

Horizon 2 (2023-2024)

Monitoring and analyzing ethical initiatives and impacts:

Launching a national discussion on AI ethics
Finalizing the AI Code of Ethics and Guidelines
Disseminating the AI Code of Ethics and Guidelines to all stakeholders

Key Performance Indicators

Number of benchmarks conducted
Number of studies executed
Preparation of an AI Code of Ethics/Guidelines
Number of AI ethical standards framed
Creation of a data ethics center

Source: Aini Suzana Arifin, “Strategy 1: Establishing AI Governance,” Malaysia Artificial Intelligence (AI) Roadmap, National AI Roadmap Townhall, March 15, 2021, https://airmap.my/st1/.

Implicit in the AI-Rmap’s logic for a broader AI governance framework for Malaysia is that AI will be ubiquitous in the country’s development trajectory. The AI-Rmap impresses the need for “continual oversight” over AI technologies that will permeate all aspects of human activity and productivity through the Internet of Things, Fourth Industrial Revolution, big data analytics, and security and surveillance. AI governance is therefore needed to set the “parameters and agencies of Artificial Intelligence in public service delivery as well as its proper role, execution, and regulation across the environment, digital commons, society, and technology.” Somewhat confusingly, the AI-Rmap defines AI governance as encompassing the “vehicular, structural, and actionable aspects” of AI in “quadruple helix society,” built upon four axes of human-centricity, explainability, transparency, and ethics.²⁴⁰ While the latter terms are commonly used in AI guidelines around the world, unfortunately, there appears to be no accompanying clarification in the AI-Rmap of how they might translate in the Malaysian context.

The AI-Rmap is an extension of the Malaysia Digital Economy Corporation’s (MDEC) still unreleased National AI Framework (NAIF), which is said to set out 20 initiatives within six key building blocks and five goals related to the economy, government, and industry, as well as people and society. NAIF aside, the AI-Rmap also takes into account seven other AI-related documents, including national policies that incorporate the development and implementation of AI. These are the Shared Prosperity Vision (SPV) 2030 (which replaces the earlier Vision 2020); the Ministry of Science, Technology and Innovation’s (MOSTI) National Science, Technology and Innovation Policy (NSTIP) 2021-2030 and 10-10 MySTIE Framework; the Ministry of International Trade and Industry’s (MITI) Industry4WRD: National Policy on Industry 4.0; the Academy of Sciences Malaysia’s Envisioning Malaysia 2050 Foresight Narrative; Malaysia Digital Economy Blueprint; and Malaysia AI Blueprint 2019.

It is worth noting that most of these documents either at least cursorily refer to, or discuss, ethics in specific connection to AI or, more generally, to technology. The Blueprint, in particular—as the Malaysia chapter on data elaborates—identifies ethics in the use of data and digital tools as one of its three guiding principles.

Usage and Impact

A common thread running through Malaysia’s boggling number of technology-related policies is the prioritization of national economic development and the modernization of the public service in order to facilitate that. Tellingly, the NSTIP itself states that its rationale is to strengthen the position of science, technology, and innovation “in the development and growth of an innovation-based economy.” Accordingly, the NSTIP introduced the acronym, “STIE”—science, technology, innovation, and economy—that is used throughout the document “to support economic growth” and to become a “high-tech nation.”²⁴¹

There is, however, an appreciation reflected in these many policies that achieving high-tech status and improving the quality of life through AI and big data analytics will have to be moored to principles of inclusiveness, economic justice, social equity, and sustainability.

Malaysia’s National Fourth Industrial Revolution (4IR) Policy, which addresses digital, physical, and biological advancement, as well as the overlaps between them, aims to facilitate socio-economic transformation through the “ethical use of 4IR technologies.” These technologies include artificial intelligence and big data in the digital domain, autonomous vehicles and 3D printing in the physical realm, and bioprinting and genetics in the biological sphere.²⁴²

The Envisioning Malaysia 2050 document acknowledges that sustainability “must also extend to the notions of harmony and prosperity through smart management of its resources and leveraging on opportunities enabled by the knowledge of its people.”

At the government level, the foray by Putrajaya (Malaysia’s administrative capital) into big data analytics as a foundation for AI applications started over a decade ago, with projects such as the Ministry of Health’s (MOH) Malaysian Health Data Warehouse (MyHDW) in 2010. Utilizing big data analytics, AI, and geographic information system (GIS), MyHDW is meant to process huge volumes of structured and unstructured data for health data management, publication, and dissemination, as well as for the development of health informatics standards. Over 90 percent of the technology is homegrown and built by the Malaysian Institute of Microelectronic Systems (MIMOS), the country’s national applied research and development agency under MOSTI.²⁴³

To preserve the privacy and security of information collected, stored, and analyzed through MyHDW, MIMOS assures the system complies with all relevant provisions under Malaysia’s Personal Data Protection Act (PDPA) 2010. There is also data pseudonymization (the practice of replacing clear identifiers such as names with pseudonyms) to conceal patient information, as well as multi-factor user authentication, in accordance with the Malaysian Administrative Modernisation and Management Planning Unit’s (MAMPU) guidelines. Additionally, a Data and Information Governance Committee manages data ownership, usage, and quality.²⁴⁴

To be sure, these are progressive measures. Privacy-enhancing technologies like pseudonymization and anonymization, in particular, are seen to offer a balance between preserving individual privacy, on the one hand, and optimizing information system functionality, on the other.²⁴⁵ They are also in line with international standards such as ISO 25237:2017 for the privacy and protection of personal health information.²⁴⁶

However, as discussed in the Malaysia chapter on data, the government is presently excluded from the PDPA’s ambit, which is a particular omission given the involvement of MOH, MIMOS, and government hospitals. There is also ambiguity about whether pharmaceutical or insurance companies have access to the database, whether access is granted by a fee, or whether the information accessed will be used for purposes other than research.²⁴⁷

A few years after the rollout of MyHDW, the government announced a broader Public Sector Big Data Analytics Pilot Project. This initiative, announced in 2013, began in 2015 as a collaboration among three agencies—the Ministry of Communications and Multimedia, MAMPU, and MDEC—with proof-of-concept implementation in sentiment analysis, price monitoring, public health, and crime prevention.²⁴⁸

With the launch of SPV 2030, Malaysia’s big data initiatives have become even more ambitious. A planned National Big Data Analytics Center (NBDAC) will enable public decision-making to be based on data analytics, in line with Malaysia’s digital government and Fourth Industrial Revolution (4IR) strategies.²⁴⁹ The NBDAC will encompass projects such as the Department of Statistics Malaysia’s (DOSM) estimation of land area and productivity in rubber plantation through satellite imagery, machine learning, and mobile positioning data. The utility of an NBDAC was also underlined during COVID-19, as it would have enabled a clearer picture of labor disruptions and the actual size of the informal work sector as Malaysia grappled with waves of movement control orders.

The AI-Rmap document sketches five national AI projects in healthcare, agriculture, education, smart city transportation, and public service. Catalyzed by the pandemic, the project on healthcare would rely on machine and deep reinforcement learning to establish an autonomous vaccine distribution and management system.²⁵⁰ Using similar AI technologies, the project on agriculture envisions an AI-driven supply chain management system for Malaysia’s important palm oil industry.²⁵¹

The project on education would utilize AI technologies such as machine learning, neural networks, and natural language processing to develop a personalized learning system equipped with automated assessment to be deployed at scale. The hope for this particular project is to match industry demand with suitable graduates and job seekers for a “future-driven workforce.”²⁵²

For smart city transportation, machine learning, big data, optimization, IoT, and blockchain would drive autonomous maintenance, repair, and operation processes for a more effective and reliable public commuter service.²⁵³

In the public service, intelligent automation through chatbots could improve process efficiency and service delivery at the federal and state government levels.²⁵⁴ Additionally, MAMPU is studying the use of AI-based facial recognition to monitor employee attendance.²⁵⁵

With the exception of the AI-in-education project, the other nationwide initiatives will center on reducing operational friction and increasing productivity. Policies and guidelines for the public sector’s use of AI and blockchain will set the parameters for ethical and effective implementation of these technologies. However, there are potential minefields even with seemingly mundane processes in innocuous sectors such as agriculture, to be discussed in the next section. Guidelines will have to be scrutinized by more than the AI-Rmap’s quadruple-helix structure of stakeholders to check for bias and to ensure ethical concepts and practices are upheld, from system design to deployment.

Case Study

Use of AI in the Judiciary

Malaysia has begun piloting AI sentencing tools in two states, Sabah and Sarawak. The impetus behind the push to utilize AI in the judicial system is to achieve greater consistency in sentencing, and to allow the courts to clear case backlogs efficiently, preventing stressful and lengthy legal proceedings. The AI tool is currently being trialed in two types of offences: drug possession under Section 12(2) of the Dangerous Drug Act and rape under Section 376(1) of the Penal Code.

The AI algorithm analyzes cases under both offences in Sabah and Sarawak between 2014 and 2019. A model is created from past case patterns, then applied to present-day cases before producing sentencing recommendations that judges can choose to adopt or deviate from.

Critics, however, warn of the risks involved when utilizing AI to make such decisions. These include the amplification of bias against minorities and marginalized groups as well as the lack of ability to consider mitigating factors and circumstances. Malaysia’s Bar Council has also questioned the validity and transparency of the algorithm, given that the training dataset used was limited to only a five-year period. In response, the Sabah and Sarawak courts, along with the software developer Sarawak Information Systems Sdn. Bhd., attempted to mitigate the risk of bias by removing the “race” variable for future sentencing guidance.

The government is also partnering with the private sector to implement large-scale projects such as urban planning and development. Alibaba’s City Brain program, which started out in the Chinese city of Hangzhou, was rolled out in Malaysia’s capital, Kuala Lumpur, as a partnership involving the tech giant, MDEC, and City Hall. In the first phase, inputs from 382 cameras and 281 traffic light junctions around central Kuala Lumpur fed data for real-time analysis and improved traffic forecasting. The project is intended to support Malaysia’s digital transformation with Alibaba’s cloud and AI software, although the data will be owned by the city.²⁵⁶

However, as others have pointed out, the fact that City Brain will also be offered as an open innovation platform to enterprises, startups, and research organizations raises privacy and surveillance concerns, since data that is initially gathered and algorithmically processed for public use may end up being utilized for profiling and commercial purposes in the future.²⁵⁷

Challenges and prospects

A 2021 study on Malaysia’s progress in big data analytics and AI proposed a national AI ethical framework to guide industry in developing and deploying AI that would be “transparent, unbiased, and beneficial to as many people as possible.” The idea, of course, is not new, given Putrajaya’s own cognizance of the imperative to integrate ethics into AI use through documents such as the AI-Rmap and the Blueprint. However, the study makes it clear that such a framework should go beyond the normative instruments that regional and international organizations, as well as multinational tech companies, have already tabled. Rather, an AI ethical framework for Malaysia should include additional principles “unique to its national values and aspirations.”²⁵⁸

This approach makes sense since even basal, global standards may need to be adapted to, or supplemented by, those that are more contextually applicable to a particular nation. In Malaysia’s case, these should address the additional challenges of government data classification, labor disruption, and inclusivity, in particular.

Government data classification

Open government data is a good practice in the digital age of making data produced or collected by governments available for free public use. It promotes good governance principles of transparency and accountability, especially if, as in Malaysia’s case, government is currently excluded from data protection legislation. Although tensions between open government data and classified government data exist everywhere, they can be complicated by a vague classification policy that might inadvertently result in an overly cautious bureaucracy being hesitant to publish information. This, in turn, could negatively impact the quality of AI data or datasets and produce automated decisions based on incomplete information that would likely not even be verifiable.

In 2021, Putrajaya took a step closer to adopting a Cloud First Policy with a collaboration between the government’s cloud service provider, MyGovCloud@PDSA, and four commercial providers: Microsoft Azure, Google Cloud, TM Cloud Alpha, and Amazon Web Services.²⁵⁹ The decision was meant to leverage the potential of data analytics, AI, and other emerging technologies.

Yet efforts have been slow to get underway. In 2019, an MDEC engagement session with 29 government agencies revealed several concerns over cloud adoption, primarily related to security. This was notwithstanding a mandate by MAMPU dating back to 2003 directing the observance of a comprehensive set of minimum technical standards covering data integration, information access, and metadata, among others.²⁶⁰

The lack of clarity on government data classification in a cloud environment was also listed as a significant concern during that session. The Chief Government Security Office (CGSO) released data classification guidance for public agencies in 2021 so that data can be tagged, categorized, and organized for more efficient, automated processing.²⁶¹ Yet informants pointed out that the data classification initiative has so far been driven more by the private rather than the public sector. Additionally, the lack of an integrated database in many government agencies, as well as varying administrative procedures, such as the imposition of a fee in exchange for data even between agencies, hamper data-sharing and classification efforts.²⁶²

Determining the parameters of what data should or should not be shared goes beyond improving bureaucratic functionality. It has substantial implications for how automated decisions are made and how the results affect people’s lives.

Labor disruption

A 2017 study calculated that 54 percent of jobs across all major economic sectors in Malaysia could be at high risk of being technologically displaced over the next 20 years. Of that figure, more than 70 percent would be semi-skilled and 80 percent would be low-skilled jobs. Women in clerical positions would be disproportionately affected, but men in more labor-intensive industries such as agriculture, mining, and construction would be overwhelmingly at risk from automation.²⁶³ These projections accord with other studies, with one estimate showing about 50 percent of work time in Malaysia being spent on highly automatable activities.²⁶⁴

Pre-emptive and mitigation strategies through technical vocational education and training (TVET), as well as reskilling programs, are already underway.²⁶⁵ There is also the anticipation that there will be new demand for labor, notwithstanding the expected increase of automation and use of AI in both the public and private sectors. Factors such as the surprising growth of Malaysia’s digital economy, despite the economic shocks of the pandemic, as well as greater investment in renewables, could create more jobs in different areas.²⁶⁶

Effectively future-proofing Malaysia’s workforce should go beyond worker retraining or reskilling in digital technologies. Jobs requiring a high level of social, emotional, linguistic, and cognitive skills will help diversify workers’ options. Some allowance has already been made for this in the Blueprint. The document outlines the need to develop creative thinking among students for a “competent and agile workforce.” It also spells out a specific initiative to nurture talent in the arts, entertainment, and recreation industries to enhance the nation’s innovation and export of digital content.²⁶⁷

Yet, as the Malaysia Education Blueprint 2013-2025 itself implicitly acknowledges in its rationale, the country’s education system has long been plagued by the rigidity of rote-learning; an exam-oriented approach; and the fragmentation of school types, from the national and vernacular to the international and religious.²⁶⁸ This last challenge has deepened divisions among Malaysia’s next generation along socio-economic and rural/urban lines, even as it has provided the curricular flexibility reflective of the nation’s diversity. This could have lasting effects on the country’s AI-powered future in two ways—by under-preparing Malaysians for a disrupted labor market, and undermining the desired transformative and inclusive impact of technology for the whole nation.

There is, however, a grimmer aspect to labor disruption that should be considered, particularly if Malaysia’s national AI project in agriculture is to focus on palm oil. The country is the world’s second largest exporter of the commodity. Automating harvesting, extraction, and supply chain management to “sustain the productivity and revenue of the plantation companies despite shortage of labor” would, on the surface, be a sound use of machine learning, deep learning, and computer vision.²⁶⁹

Given the controversy surrounding allegations of forced, undocumented labor in Malaysia’s palm oil plantations, the integration of AI in these production processes could well alleviate some of the industry’s problems in this regard. Yet, as examples elsewhere have shown, the use of AI technologies can also strengthen power and control in the hands of corporations to the detriment of mainly low-wage workers. This, in turn, can exacerbate income inequality or worse, exploitation.²⁷⁰ Any code of ethics guiding the use of AI in this and, indeed, other industries with a stark power imbalance would have to make allowances for both the foreseeable and unintended consequences of reproducing pre-existing inequities through automation.

Inclusivity

One of the most discomfiting risks of AI in a multi-ethnic and multi-religious society is its proven ability to propagate bias and discrimination through opaque algorithms. In the context of Malaysia, two related and sensitive questions arise. First, how will the nation’s delicate balance of communal relations play into and affect algorithmic datasets without worsening real-world biases? Second, big data initiatives notwithstanding, with gaps in the availability and quality of data, can Malaysia credibly train AI systems, and can these technologies redress or reverse latent policy biases?²⁷¹

Data, after all, is a reflection of historical, political, and social realities. The ethnic, religious, and linguistic cleavages that post-colonial Malaysia inherited at independence continue to underpin public and private debates on education, the legal and justice systems, and economic policies. If anything, these frictions have only been amplified with greater political space and the advent of social media since the early 2000s. Not addressing these communal fractures in real life and delegating decisions to algorithms, instead, will only embed prejudice under the guise of technological neutrality.

One way to assess whether AI-powered systems will serve Malaysia’s diverse population equitably is to examine the ethnic or national-origin composition of the scientific research community in the country. This shines a light into how welcoming it is to different perspectives, which could in turn stimulate higher-quality research.

A study in 2016 concluded that Malays greatly increased their research participation and publication in Malaysia over three decades, from 20 percent in 1982–1984 to 65 percent in 2012–2014. There was a corresponding decline among Malaysian Chinese and Indian authors (although their absolute numbers increased because of a rise in Malaysian scientific output in total). Malay researchers demonstrated particular strength in engineering and technology, as well as in physics, and doubled their representation in clinical medicine, which is a field traditionally dominated by Indians.

Unsurprisingly, research and writing collaborations with foreign scientists strongly reflected the ethnic divisions in Malaysia; that is, Malays (who are also Muslim, as defined by Malaysia’s constitution) favored working with counterparts from Muslim-majority countries, Indians with colleagues from India, and Chinese with those from China. The expansion in scientific output among Malays is remarkable when compared against the steadier population growth of Malays, as a collective group, in the same time period.²⁷²

These figures on science are testament to the country’s affirmative action agenda dating back to the New Economic Policy of 1970 to correct for ethnic-wealth disparities. However, this approach, which started out with good intentions to uplift Malaysia’s bumiputera (sons of the soil) population, has proven polarizing over the years.²⁷³

The country’s orang asal or orang asli (indigenous community) are categorized as bumiputera, yet they continue to be hardest hit by poverty. Official data from 2010 showed that over 50 percent of the community in peninsular Malaysia was categorized as poor and 33 percent as hard-core poor. By comparison, the national average of hard-core poor was 0.7 percent.²⁷⁴

The orang asli in peninsular Malaysia were not actually included in DOSM’s 2016 household income survey, even though they numbered around 178,000 people. By contrast, 80,000 households were surveyed. DOSM explained that the Department of Orang Asli Development (JAKOA) had those detailed statistics, but did not provide a reason why they had been left out of the survey itself.²⁷⁵ As it is, observers point to the challenge of data exclusion in countries such as Malaysia as a larger concern than data protection and privacy.²⁷⁶ The lack of quality data, as well as decentralized access to datasets, could degrade the algorithmic training process and consequently the value of decisions derived from that.

Like most indigenous communities elsewhere, Malaysia’s orang asal/orang asli provided for themselves by foraging, hunting, and subsistence farming on native customary land. However, repeated infringements of their land rights through logging, mining, and cash crop plantations have left them marginalized, and plagued by malnutrition and lack of access to basic services. Although the government established JAKOA as a dedicated agency for indigenous affairs, scholars and non-governmental organizations argue that its approach has been to assimilate and resettle the population in the name of sustainable development.

In the past, the government attributed the continued impoverishment of the orang asal/orang asli to their failure to view development through the same national lens.²⁷⁷ This perspective questions the meaning of references to harmony, prosperity, sustainability, and progressiveness in the Envisioning Malaysia 2050 document, which also talks about “endogenous STI” and knowledge of the country’s people. Presumably, this would also encompass indigenous knowledge among Malaysia’s native communities in peninsular and east Malaysia with their deep connections to nature and biodiversity.

Another indication of inclusivity in advancing an ethical AI ecosystem is female participation in STEM subjects. In Malaysia, women fall behind men at both the tertiary and professional levels. Of the number of STEM tertiary graduates in 2018, 34.2 percent constituted women compared to 65.8 percent for men.²⁷⁸

Interestingly, in the early 2000s, as the number of women in higher computer science education decreased in many Western countries due to perceptions of the field as “masculine,” Malaysia presented a different picture. At the time, women constituted about half of all students and the majority of the faculty in computer science and information technology (IT).²⁷⁹ One study in 2006 concluded there was no gender bias with regard to how those subjects were viewed by female students. In fact, when compared to their male counterparts, female students were more certain that they would go on to pursue a career in computing or IT.²⁸⁰

Yet the overall picture is a little more complex. Malaysia has had a national policy on women since 1989 to achieve gender equality, yet has the third lowest female labor force participation rate in the ASEAN region.²⁸¹ At the same time, the country has a reverse gender gap in school enrollment at all levels, with males being significantly underrepresented in public university enrolment in every field of study except engineering, manufacturing, and construction. There is also a trend of male underperformance at the secondary education level.²⁸² So, as Malaysia works toward bringing on board more women into STEM for more ethical AI systems, society should also ensure that it doubles down on engaging males in education so as not to create a cohort of “lost boys”, or educationally marginalized young men.²⁸³

Conclusion

Malaysia’s AI-Rmap document, with its proposal for an AI Code of Ethics, and pending National AI Framework are encouraging steps to integrate ethical principles and standards into the country’s transition to AI technologies. A suite of other complementary digital strategies, policies, and blueprints will help institutionalize this approach for the long term.

However, it is still a huge leap from this point to what was initially envisioned with the birth of the Multimedia Super Corridor (MSC), which set the stage for nearly all of Malaysia’s digital ambitions and policies—a networked ecosystem of ICT- and IT-enabled industries that would “develop new codes of ethics in a shrunken world,” “set global standards in flagship applications,” and facilitate a “world-leading” and “harmonized global framework of cyberlaws.”²⁸⁴

While Malaysia has been a member of the International Telecommunication Union (ITU) since 1958 and the Department of Standards Malaysia is a member of both the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC), it does not currently participate in the ISO and IEC Joint Technical Committee (JTC) on Artificial Intelligence. The JTC serves as the focal point for the Committee’s standardization program on AI, including on ethics, and provides guidance on developing AI applications. Drawing up new codes of ethics and international standards requires consistent presence and participation in these sorts of forums.

But it would have to start with multi-stakeholder discussions at the national level. This would require delicate conversations about the types of values and principles underpinning the multiculturalism of Malaysia, which could then be translated into an AI Code of Ethics to govern technology. It would also warrant a recap of the nine challenges first laid out in Vision 2020 and recalled by the Envisioning Malaysia 2050 document as a recurring theme of “a united, democratic, moral and ethical, liberal and tolerant, just and equitable, prosperous, scientific and progressive society.”²⁸⁵ To paraphrase an informant, ethical AI in a nation of diversity can only be realized if there is first an embrace of that diversity. The proof will be in the coding.

Endnotes

²⁴⁰ Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁴¹ “National Science, Technology and Innovation Policy (NSTIP) 2021-2030,” Ministry of Science, Technology and Innovation, February 2022, 2.

²⁴² “National Fourth Industrial Revolution (4IR) Policy,” Economic Planning Unit, June 29, 2021.

²⁴³ “TM National Health Data Warehouse (MyHDW),“ Mimos, 2019.

²⁴⁴ Ibid.

²⁴⁵ Stakeholder consultation; Yves-Alexandre de Montjoye, Ali Farzanehfar, Julien Hendrickx and Luc Rocher, “Solving Artificial Intelligence’s Privacy Problem,” Artificial Intelligence and Robotics in the City 17 (2017), 80-83.

²⁴⁶ “ISO 25237:2017 Health Informatics – Pseudonymization,” International Organization for Standardization, January 2017.

²⁴⁷ Shazwan Mustafa Kamal, “Big data in healthcare: What we (need to) know,” Malay Mail, April 21, 2017.

²⁴⁸ “Public Sector Big Data Analytics (DRSA),” My Government – the Government of Malaysia’s Official ortal, accessed June 6, 2022.

²⁴⁹ Ibid.

²⁵⁰ “Healthcare,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁵¹ “Agriculture,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁵² “Education,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁵³ “Smart Cities Transport,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁵⁴ “Public Services,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁵⁵ Shirley Tay, “Cloud, AI and Blockchain: Inside Malaysia’s digitalization strategy,” Gov Insider, April 22, 2021.

²⁵⁶ “Kuala Lumpur to build ‘City Brain’ with Alibaba Cloud,” International Telecommunication Union, April 7, 2020.

²⁵⁷ Abigail Beall, “In China, Alibaba’s data-hungry AI is controlling (and watching) cities,” Wired, May 30, 2018.

²⁵⁸ “Malaysia AI Blueprint 2021 Annual Report,” Bigit, 2021.

²⁵⁹ “Government Cloud Services,” My Government - the government of Malaysia’s official portal, accessed June 5, 2022.

²⁶⁰ “Ir Wan Murdani, “CSA APAC Congress 2019: An Update on the Malaysian Cybersecurity & Cloud Security Landscape,” Malaysia Digital Economy Corporation, 2019.

²⁶¹ “Soalan Lazim Mengenai Klasifikasi Maklumat/Data Dalam Penggunaan Pengkomputeran Awan” [Frequently Asked Questions on the Classification of Information/Data in the Use of Cloud Computing], Cloud Computing Committee, August 2021.

²⁶² Ahmad Ashraf Ahmad Shaharudin, “Open Government Data in Malaysia: Landscape, Challenges and Aspirations,” Khazanah Research Institute 3/21, April 15, 2021; Stakeholder consultation.

²⁶³ Allen Ng, “The Times They Are A-Changin’: Technology, Employment, and the Malaysian Economy,” Khazanah Research Institute, April 28, 2017.

²⁶⁴ Ee Huei Koh and Nimal Manuel, “Automation and Adaptability: How Malaysia Can Navigate the Future of Work,” McKinsey, February 17, 2020.

²⁶⁵ “TVET Country Profile: Malaysia,” UNESCO-UNEVOC TVET Country Profiles, June 2019.

²⁶⁶ “Malaysia e-commerce income soared 17.1 per cent to RM279.0 billion in the third quarter 2021,” Department of Statistics Malaysia Official Portal, November 10, 2021.

²⁶⁷ Malaysia Digital Economy Blueprint, (Putrajaya, Malaysia: Economic Planning Unit 2021), 81, 88.

²⁶⁸ Malaysia Education Blueprint 2013-2025 (Preschool to Post-Secondary Education, (Putrajaya Malaysia: Ministry of Education Malaysia, 2013). See also Chan Soon Seng, Noor Azimah Abd Rahim, and Nina Adlan Disney, “Malaysia’s Education Challenges #1: Our Education System: Overview of Challenges and Solutions,” BFM: The Business Station, November 15, 2021.

²⁶⁹ ”Agriculture,” Malaysia Artificial Intelligence (AI) Roadmap, Ministry of Science, Technology and Innovation, accessed May 30, 2022.

²⁷⁰ Andie Burjek, “The ethical use of AI on low-wage workers,” Workforce.com, March 9, 2020; Daron Acemoglu and Pascual Restrepo, “Unpacking Skill Bias: Automation and New Tasks,” National Bureau of Economic Research, January 2020; Kate Crawford, et al, “AI Now 2019 Report,” New York: AI Now Institute, 2019.

²⁷¹ Jun-E Tan, “Human Rights Concerns on Artificial Intelligence in Southeast Asia: An Overview,” Coconet, December 24, 2019; “IGF 2017 WS #303 Artificial Intelligence in Asia: What’s Similar, What’s Different? Findings from our AI workshops,” Twelfth Annual Meeting of the Internet Governance Forum, December 20, 2017.

²⁷² Grant Lewison, et al, “The Contribution of Ethnic Groups to Malaysian Scientific Output, 1982-2014, and the Effects of the New Economic Policy,” Scientometrics 109, no. 3 (2016).

²⁷³ Jomo K.S., “The New Economic Policy and Interethnic Relations in Malaysia,” Identities, Conflict and Cohesion Programme paper number 7, United Nations Research Institute for Social Development, September 2004.

²⁷⁴ Ooi Kiah Hui, “Poverty, Inequality and the Lack of Basic Rights Experienced by the Orang Asli in Malaysia,” Submission on Malaysia in advance of country visit of the Special Rapporteur on Extreme Poverty and Human Rights OHCHR, United Nations Human Rights Office of the High Commissioner, accessed June 5, 2022.

²⁷⁵ “Orang Asli not part of income survey,” The Star, September 2, 2019.

²⁷⁶ Jun-E Tan, “Human Rights Concerns on Artificial Intelligence in Southeast Asia: An Overview.”

²⁷⁷ Md. Khaled Saifullah, Muhammad Mehedi Masud, and Fatimah Binti Kari, “Vulnerability context and well-being factors of indigenous community development: a study of Peninsular Malaysia,” AlterNative: An International Journal of Indigenous Peoples 17, no. 1 (February 23, 2021), 94-105.

²⁷⁸ “Gender Statistics,” World Bank, accessed June 5, 2022.

²⁷⁹ Lagesen, Vivian Anette. “A Cyberfeminist Utopia?: Perceptions of Gender and Computer Science among Malaysian Women Computer Science Students and Faculty.” Science, Technology, & Human Values 33, no. 1 (January 2008): 5–27.

²⁸⁰ Mazliza Othman and Rodziah Latih, “Women in Computer Science: No Shortage Here!,” Communications of the ACM 49, no. 3 (March 2006), 111-114.

²⁸¹ Official Portal of the Department of Women’s Development, accessed June 5, 2022; “Toward Better Economic Opportunities for Women: Lessons from Malaysia,” World Bank, April 21, 2020.

²⁸² Jonathan Yong, “The Gender Gap in Malaysian Public Universities: Examining the 'Lost Boys',” Journal of International and Comparative Education 6, no. 1 (April 2017), 1-16, doi: 10.14425/JICE.2017.6.1.0116.

²⁸³ Ministry of Education Malaysia, Malaysia Education Blueprint 2013-2025.

²⁸⁴ Mahathir bin Mohamad, “The Opening of Multimedia Asia on Multimedia Super Corridor,” Transcript of speech delivered at Putra World Trade Centre, Kuala Lumpur, August 1, 1996; “MSC Malaysia Status,” Malaysian Investment Development Authority, accessed June 5, 2022; Adrian Regan, “The Multimedia Super Corridor,” World Trade Organization Information technology Symposium, July 16, 1999.

²⁸⁵ Mahathir Mohamad, Malaysia: The Way Forward (Vision 2020), Kuala Lumpur: Percetakan Nasional Berhad, 1991; Envisioning Malaysia 2050: A Foresight Narrative (Kuala Lumpur: Academy of Sciences Malaysia, 2017), 59.

Download Report ↓