ASEAN unveiled a framework on AI regulation on Friday, however, experts warned that this could pose some challenges as member states are at different stages of digital development.
Singapore's Ministry of Communications and Information (MCI) said, "The objective of the guide is to establish common principles for trusted AI and suggest best practices on how to implement trusted AI in countries across Southeast Asia."
Singapore is hosting the fourth Southeast Asia Digital Ministers' Conference, scheduled for Feb. 1-2, where emerging digital issues, including artificial intelligence and cyber fraud, will be discussed.
Kristina Fong, principal researcher for economic affairs at the Center for Southeast Asian Studies, said the guide takes a "light-touch, flexible" approach to the risks posed by AI, reflecting the challenges posed by development gaps among Southeast Asian member countries.
Fong noted that there are differences between countries, "not only in terms of digital capabilities, but also in terms of the maturity of regulators, institutional capacity, and even at the level of the rule of law."
"That's why we need a platform to discuss these risks and issues, to be able to deal with them in a coordinated manner, and more importantly, to figure out a way forward for Southeast Asian member countries that are currently lagging behind." Fong said as follows.
The ASEAN Guidelines on Governance and Ethics of Artificial Intelligence (AI) sets out a number of national and regional level recommendations that can be used by governments to inform the responsible design, development and deployment of AI systems.
ASEAN consists of Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand and Vietnam.
Singapore leads the way
The release of OpenAI's ChatGPT in November 2022 led to an instant spike in interest in AI, as the chatbot became instantly popular for its ability to generate responses similar to those of humans.
"Singapore was one of the first countries to release a national AI strategy," said Josephine Teo, Singapore's minister of communications and information, in 2019, adding that Singapore wants to "maximize the use of artificial intelligence." Singapore released a new National AI Strategy 2.0 in December last year.
Singapore leads the Asia-Pacific AI Readiness Index 2023 released by Salesforce, which assesses 12 countries. Other ASEAN member countries, such as Malaysia, Indonesia, Vietnam, the Philippines, and Thailand, placed lower in the rankings, in positions eight through twelve.
"The challenge for a region like ASEAN is that different countries in the region are at different stages of digital development. This means that countries are facing very different policy issues and thinking from very different perspectives," said Kenddrick Chan, senior policy analyst at the Tony Blair Institute for Global Change.
"Even on some of the basic regulations set to effectively manage AI risks, such as cybersecurity and personal data protection, countries have not agreed on the progress of implementation or the effectiveness of regulation," said Fong of the ASEAN Research Center.
In the ASEAN guide, country-level recommendations include developing AI talent, upgrading employee skills, and investing in AI research and development. Regional-level recommendations include the establishment of a working group to advance the recommendations and collect implementation cases.
The guide provides members with case studies of Asian companies and public sectors that have implemented AI governance measures in AI design, development and deployment, such as Gojek and Singapore's Smart Nation Group.
"It is clear that the private sector will play a crucial role in AI governance. Therefore, the government should seek mechanisms to promote public-private partnerships, both at the national and regional levels," Chen said.
"This will mean that we will have better policy development, implementation and regulation. So the ASEAN AI guidelines are a good start, but as the AI field continues to evolve, the guidelines and regulations will have to be updated as well," Chan said.
Related articles
DeepSeek-R1-Lite, a homegrown inference model comparable to o1-preview, is online!
Copilot Agents Intelligent Body Agents make their grand entrance!
Copilot Agents Workplace Encounters: the CEO's New Favorites
Copilot for PowerPoint has undergone major changes, and these are the key points that have to be looked at: rewriting, translating, illustrating, annotating
Le Chat User Interface Upgraded, Mistral Officially Enters the Consumer Market to Compete with ChatGPT