AI-Pac July edition: the Flint digest on AI policy in Asia-Pacific

AI-Pac is Flint’s new digest providing our take on the major developments in Asia-Pacific AI policy and regulation. In a fast-moving environment, this newsletter is designed to help you stay on top of the regulatory agenda and to consider what it means for your business. Please see here for our first four blogs, covering January to June. In each update, we will set out:

• Major regulatory developments across Asia-Pacific regarding AI regulation;
• Overall trends of AI regulation in Asia-Pacific;
• What this means for business and how business might best prepare.

If you would like to discuss how AI policy and regulation in Asia-Pacific will impact on your business, please get in touch with David and Ewan here.

July was a bumper month for AI policy as several Asia-Pacific governments advanced domestic policy initiatives and sought to influence the global debate. Key developments included: 

• China’s government presided over the annual World AI Conference in Shanghai, championing an alternative approach to Western-led global governance on AI. 
• Vietnam published a draft law on regulation of digital technologies and AI that included elements of both Chinese- and EU-style AI regulation. 
• Taiwan published a draft law on AI regulation as part of a wider approach weighted towards promoting AI adoption over implementing strict rules.
• The Philippines outlined an AI roadmap, aiming to accelerate AI adoption that addresses real-world economic and social challenges in the country.
• Japan’s competition regulator outlined its interest in the competition risks relating to generative AI, following in the footsteps of India and Korea. 

China: pursuing its own vision of AI global governance

From 4 to 6 July, the annual World AI Conference (WAIC) was held in Shanghai. Held since 2017, the WAIC brought together senior Chinese government officials, diplomats from China-friendly countries such as Belarus, and industry figures - including Elon Musk - to discuss the potential of AI innovation and the importance of effective governance. 

A key theme at the Conference was the importance of a globally inclusive approach to AI development and governance. China’s government used the Conference to release the Shanghai Declaration on Global AI Governance. This calls for global cooperation in promoting AI development and for the UN to take a leading role in global AI governance. This followed on from the adoption - with US support - at the UN General Assembly of a China-led resolution that called for a “free, open, inclusive and non-discriminatory” environment.

These developments reflect China’s ambitions to position itself as a global leader on AI and a partner to developing nations in their ambitions to digitalise and drive AI adoption. China’s emphasis on a non-discriminatory approach to AI development seeks to draw a contrast with US export and investment controls on AI and other sensitive technologies. Its call for the UN to take the lead on AI governance aims to present an alternative to existing Western-led global initiatives on AI governance, such as the AI summits in the UK and Korea and the G7 Hiroshima Principles. China has not yet progressed from rhetoric to substance when it comes to progressing meaningful global initiatives on AI governance. Nevertheless, its ambitions underline the fragility of global cooperation on AI safety and the high risk of greater fragmentation in the coming years.

Vietnam: combining Chinese- and EU-influences approaches to AI regulation

On 2 July, Vietnam’s government published the draft “Law on Digital Technology Industry” aimed at facilitating and regulating emerging digital technologies, including AI. The draft law is subject to public consultation until 2 September. 

The draft law is set within the context of Vietnam’s broader National Strategy on R&D and Application of AI, which outlines its ambitions to become a hub for AI innovation in Southeast Asia and globally by 2030. In January, Vietnamese conglomerate, Vingroup, launched ViGPT, Vietnam’s first publicly available homegrown generative AI model. An important consideration for Vietnam is to develop domestic AI capabilities so as to mitigate the risk of dependence on foreign AI developers.

The draft law’s AI-focused provisions reflect the influence of both EU and Chinese approaches. On the one hand, the draft formally lays out a risk-based regulatory approach to AI systems, likely inspired by the EU AI Act. On the other hand, similar to China’s Interim AI Measures on generative AI, it bans the exploitation of digital technology to inflict harm upon state interests or to disrupt social order. The draft law’s approach is distinctive compared to other countries in the region - for instance Singapore’s more flexible, co-regulatory approach - indicating that businesses operating across southeast Asia are facing an increasingly fragmented regulatory environment.

The draft law outlines a framework, but with further detail to come on the classification of risk and on the obligations of providers. For AI developers, there is the risk that Vietnam’s approach draws further on the precedent established by China, including the threat of banning generative AI models whose outputs fail to conform to the ideology and political sensitivities of the governing authorities. Mitigating against this will be Vietnam’s recognition of the importance of multinational investors and developers in realising its ambitions for its digital economy. 

The draft Law is scheduled to be submitted to the National Assembly for its review and input in the October 2024 session and approval in the May 2025 session, though this may be delayed given uncertainty over the succession of the late Communist Party general secretary, Nguyen PhuTrong. For businesses affected, it will be important to closely monitor the progress of the draft law and to consider engaging with the Ministry of Information and Communications on the substance.

Taiwan: emphasis on AI development while managing the risks

On 15 July, Taiwan’s government issued the draft AI Basic Act for public consultation. The consultation is open until 13 September.

Taiwan’s overarching approach towards AI has been to leverage its traditional strengths in advanced manufacturing to develop leading AI research & development capabilities and to accelerate the adoption of AI in strategic sectors. Premier, Cho Jung-tai, recently announced the government’s ambition to attract significant domestic and foreign investment into its AI industry. Military strategists are also monitoring the use of AI on the battlefield in Ukraine and considering its applicability to a potential military confrontation with mainland China. 

This emphasis on development means that the government is mindful of the risks of overregulating the technology. At the same time, it faces pressure from human rights advocates who have highlighted the particular risks AI may present for human rights, a sensitive issue in light of Taiwan’s history of authoritarian government. There are also concerns over the potential for China to seek to use AI-fuelled disinformation to destabilise Taiwan’s democracy as well as over AI-enabled fraud. 

The draft Act sets general principles and high-level tasks for the Government’s promotion of AI development, focusing on privacy protection, risk management, and international alignment. The Ministry of Digital Affairs (MODA) is tasked with developing an AI risk classification framework that aligns with international standards. 

After the consultation, the National Science and Technology Council aims to deliver the draft’s final version to the Executive Yuan (Taiwan’s top administrative body) by the end of October, after which the Executive Yuan will submit the act to Taiwan’s parliament. Affected businesses should review their engagement with both MODA and the NSTC.

Philippines: focus on economic and social potential of AI applications

On 3 July, the Philippines government launched the National Artificial Intelligence (AI) Strategy Roadmap 2.0 (NAISR 2.0) and Center for AI Research (CAIR). The NAISR 2.0 builds on the Philippines’ first AI roadmap, launched in 2021, and outlines a number of areas of strategic focus in order to drive AI innovation that improves the lives of ordinary Filipinos. 

The NAISR 2.0 seeks to address obstacles to AI adoption, including limited use cases, workforce readiness, and uncertainties surrounding AI regulation. It aims to increase the national R&D budget from 0.30% to 1%. The CAIR has been tasked with developing AI use-cases to address real-world challenges in manufacturing, agriculture, urban planning, and disaster resilience.

As part of the roadmap, the government also aims to produce a “national governance framework” that, according to officials, will establish “the approved scope and limitations of what developers and stakeholders can do”. The Department of Trade and Industry (DTI) is working with the private-sector led Analytics & Artificial Intelligence Association of the Philippines in drafting the governance framework and is aiming to publish it by the end of 2024. The framework is not expected to include prescriptive new rules on AI development but is likely to take a flexible approach aimed at providing businesses with greater clarity and non-binding guidance on how they can effectively mitigate risks relating to AI.

The roadmap underscores how governments across the region are seeking to realise the potential of AI for their economies and societies. In the Philippines and elsewhere, governments will look to businesses to contribute to policy initiatives such as AI skilling programs and to develop use-cases that can improve citizens’ wellbeing.   

Japan: competition and AI in the spotlight, again

On 9 July, Kazuyuki Furuya, chairman of the Japan Fair Trade Commission (JFTC), revealed in an interview that the JFTC plans to conduct a survey on competition risks relating to generative AI in the autumn.

Furuya described himself as open-minded, emphasising that the JFTC plans to “monitor the market without preconceived notions”. However, in a later conversation with reporters, Tetsuya Fujimoto, secretary general at the JFTC, identified concentration in advanced AI chips, cloud-computing systems, specialised personnel, and data as areas of concern.

The JFTC follows in the footsteps of regional counterparts in India and Korea in reviewing competition risks in the sector (see here for our analysis of those countries in a previous blog). The JFTC is also in communication with Western competition regulators on these issues through the G7, and will be carefully considering its position on the issues raised in the recent joint statement by the CMA, European Commission, and US FTC on competition in generative AI foundation models. 

For now, the JFTC’s messaging is more cautious than its Western counterparts. Taking the precedent of other market interventions such as ex-ante regulation of digital markets, the JFTC may wait and see how Western regulators tackle this issue rather than itself make the first move.

This blog was written by Ewan Lusty and David Skelton, who are both based in Singapore, with the input of Flint’s Senior Adviser network based across Asia-Pacific. Ewan has been supporting Flint clients on a range of digital policy issues for over four years, prior to which he advised UK government ministers on online safety regulation. David spent seven years at Google, most recently working within the Global Strategy and Innovation team, as well as leading Google’s Policy Planning team within Asia Pacific.