Towards Algorethical AI
Recently, in a report run by a local daily, the Southeast Asia Regional Centre for Counter-Terrorism (SEARCCT) drew our attention to its finding that international terror organisations have resorted to a new strategy of abusing artificial intelligence (AI) to spread their ideology and propaganda and lure new recruits.
The above move could actually be considered to be one of the many attempts at weaponising AI, something which should have been more than sufficient to alarm us about the evolving threat of AI misuse, especially if it involves what is termed State Terrorism.
Serious and long-term efforts to address such issues of AI abusage as the aforementioned require that we deal with matters pertaining to ethics and laws as well.
On the other hand, seemingly, Malaysians at large have shown their eagerness to embrace AI.
The question is, are they really ready for this state-of-the-art technology?
Like it or not, their readiness has to be fairly assessed.
It is therefore pertinent that we take a look at the AI Preparedness Index released by the International Monetary Fund (IMF) slightly more than a month ago.
The Index purportedly seeks to assess the level of AI readiness across 174 countries based on four primary foci, namely, their digital infrastructure; innovation and economic integration; human capital and labor market policies; as well as regulation and ethics.
With regard to regulation and ethics in particular, Malaysia’s score is 0.17, with Finland at the top of the ladder scoring 0.23 while the scores of our closest neighbours, Singapore is at 0.22; Indonesia, at 0.16; Thailand, at 0.13; and Brunei, at 0.12, respectively.
Together with Indonesia, Thailand and Brunei, Malaysia is placed in the emerging market economies (EME) whose overall index is 0.12, while Singapore is grouped with countries categorised as advanced economies (ADVEC) whose overall index is 0.19.
The overall score of low-income countries (LIC) in the Index is 0.08.
Insofar as the details of Malaysia’s AI readiness are concerned in terms of ethics and law, it seems that the only official master plan which we could rely on at present is the one outlined on a few pages in the almost five-year-old Malaysia AI Roadmap 2021‒2025, particularly on pages 28 to 30, as well as 33 to 39.
We thus come across thereon seven Principles of Responsible AI, namely: fairness; reliability, safety and control; privacy and security; inclusiveness; transparency; accountability; and pursuit of human benefit and happiness.
The roadmap does provide for each principle, a three-to-four-line explanation.
In fortifying such principles, we should also attend to earlier or concurrent initiatives and attempts undertaken by parties outside Malaysia who share similar concerns.
Apart from what the European Union had come up with in 2021, also relevant in this regard is the ground-breaking, multi-religious approach spearheaded in early 2023 by representatives from the three world religions, Christianity, Islam, and Judaism, with the support of Microsoft, IBM and Cisco, and in the second week of July this year in Hiroshima, joined by religious leaders representing various faiths of the Eastern world such as Buddhism, Hinduism, Zoroastrianism, Bahais and Sikhism.
Collectively, they call for ALGORETHICS, i.e. the ethical use of AI as defined by the six core principles:
First, transparency: in principle, AI systems must be explainable.
Secondly, inclusion: the needs of all human beings must be taken into consideration so that everyone can benefit and all individuals can be offered the best possible conditions to express themselves and develop.
Thirdly, responsibility: those who design and deploy the use of AI must proceed with responsibility and transparency.
Fourthly, impartiality: do not create or act according to bias, thus safeguarding fairness and human dignity.
Fifthly, reliability: AI systems must be able to work reliably.
And finally, security and privacy: AI systems must work securely and respect the privacy of users.
As regards this final principle, we should take note of an initiative currently being taken by the Government of Malaysia, in particular the Digital Ministry, aimed at establishing a Malaysian Data Commission to enable adequate frameworks and policies to be developed to protect people’s data and promote the use of data centres in the country.
As is clear, all the aforementioned principles remain general and perhaps abstract.
Unless and until they are fleshed out with operational details, they will be taken as, at best, mere noble aspirations.
Certainly, many more need to be done to address ethical and legal issues related to AI in a substantive way so that AI, however advanced it may be developed to become, shall continue to be a useful tool for the benefits of humankind.
In this regard, concerned parties are eagerly awaiting the Artificial Intelligence Governance and Ethics (AIGE) which is being developed by the Ministry of Science, Technology and Innovation (MOSTI) and the Digital Ministry and, as reported recently, is expected to be launched before the end of this year.
