Following our webinar on AI legal developments in ASEAN and China, we’ve published a series of three articles looking closely at the different countries. This second piece covers Vietnam, Thailand and the Philippines.
(The first article looks at Singapore, Indonesia and Malaysia, and the third provides an overview of developments in China.)
Vietnam
Two years on from issuing its national strategy on AI development, with a vision to 2030, Vietnam has achieved several significant milestones.
FPT Corporation – one of the largest IT-service companies in Vietnam – has now announced a strategic cooperation with Nvidia. The aim is to promote AI research and provide services for both Vietnamese and global customers. FPT plans to invest USD 200 million to build a factory with AI supercomputers operating on Nvidia’s technology.
Other domestic companies that have made significant achievements in AI research and application include VinAI, Viettel AI, and VNPT AI. These major players have also invested in building data centres – facilities that are the backbone of AI development – with Vietnam having 39 such data centres as of October 2023.
“There has not yet been any comprehensive AI legal framework in Vietnam, but the country has a clear intention to develop this in the near future. In the meantime, existing regulations in IP, IT, eCommerce, cybersecurity and data protection may be used,” explains Rouse Principal for Vietnam, Yen Vu.
“AI systems that use copyrighted works without permission of the copyright owners can constitute infringement in Vietnam. Use of the content to train AI models is unlikely to be a valid defence, even if the AI services are provided free of charge,” says Vu.
There is also the risk of AI systems overstepping Vietnam’s data-privacy rules. This may attract regulatory scrutiny when personal information is involved. Requirements to establish a local office may also come into play.
Vietnamese authorities pay close attention to cybersecurity issues and may apply a strong enforcement approach to any breaches. Information collection and sharing can be subject to criminal prosecution, especially when such actions are considered as being against the interests of the state.
Thailand
In 2022, the Thai government came out with a national AI strategy and action plan. The country’s vision is to have an effective ecosystem to promote AI development, with the aim of improving the economy and quality of life by 2027.
Thailand’s Office of the National Digital Economy and Society (ONDE) has issued AI Ethics Guidelines, developed in consultation with Microsoft’s local office. The country is taking a risk-based approach to AI regulation that has parallels with the European Union’s AI Act. There are currently two pieces of draft legislation in the works.
The first is the Draft Royal Decree on Business Operations that Use AI System, which is intended to be a general regulation for AI. It ranks AI systems into three categories: prohibited, high-risk, and limited risk.
“Different sectors may issue their own regulations later, but they must adhere to a standard that is not lower than this general regulation,” explains Rouse Principal for Thailand, Peeraya Thammasujarit.
The other draft legislation is the Draft Promotion and Support of AI Innovations Act. It aims to promote and support the AI ecosystem through mechanisms such as registration, an AI sandbox, data-sharing guidelines, contract terms and more. There is also an AI compensation fund, to provide compensation for damages from AI use where no liable party can be found.
“It’s not mandatory that companies register, but there are benefits to doing so. For example, registered businesses can request consultation from the AI Clinic, join the AI testing centre, and access data for AI training,” says Thammasujarit.
Thailand has been active in terms of data-centre development, with 13 centres currently registered in the country. Amazon Web Services (AWS), Google and Microsoft have all expressed their commitment to bring cloud services to Thailand. The government is supporting this industry through a National AI Service Platform and a Supercomputer Center (ranked as the most powerful in ASEAN in 2022).
“Despite these developments, there do not seem to be any planned changes in copyright law to accommodate data training. Thai law does not directly address data training, so the use of copyrighted work for data training may not be exempted from copyright infringement under existing copyright law. However, this position has not been tested in the courts yet,” explains Thammasujarit.
Thai copyright law defines an author as a person, which can be either a natural person or a legal person. As AI is neither of these, the scholarly thinking is that AI cannot be recognised as an author and thus the work created by AI cannot be recognised as copyright.
“This could be changed in the future, either by amending the Thai Copyright Act to recognise AI as an author, or to amend general law to recognise AI as a legal person. It remains to be seen,” observes Thammasujarit.
Philippines
The Philippines launched its AI strategy in May 2021. The aim is to position the country as a data-processing hub that provides high-value data analytics and AI services to the world.
“We believe the Philippines has an existing comparative advantage, given our strength in Business Process Management (BPO). This sector generates almost 36 billion US dollars each year – contributing about 7.5% to the country’s GDP,” explains Rouse Principal for the Philippines, Edmund Jason Baranda.
“AI has already been transforming the BPO industry here through use of the technology in our contact centres. BPO companies are at the forefront of investing in and adopting AI. More than 50 tech startups in the Philippines – many of which serve the BPO industry – are using AI as their core technology,” he says.
One of the key points in the country’s AI strategy is to establish a National Center for AI Research (N-CAIR), with the goal of turning the country into an AI powerhouse. The strategy also aims to identify key areas in both R&D and technology application, so that investment and resources can be steered towards these.
“The private sector is currently driving local R&D in AI. The Philippine government would like to see more cooperation between government, industry and academia,” explains Baranda. “It has been suggested that AI may displace a lot of workers – especially in the BPO industry – so part of the government strategy is to reskill people to limit this.”
The Philippines currently has no legislation that directly deals with or regulates AI, but there are some laws that refer to the technology:
“On the copyright side we do not have an express provision relating to copyright and AI. But we have a provision that says the author of a copyright work should be a natural person. Based on this, an AI work may not be protected by copyright,” says Baranda.
“As there’s no law – or even a bill – to regulate the use of copyrighted works for AI training data, it has been argued that this should fall under fair use. We have to wait for a decision from our Supreme Court. Our courts often cite decisions of the US Supreme Court when there is no local jurisprudence, so it may be helpful to monitor developments in the US if you want to predict how our courts will decide,” he says.
The following relevant bills are currently pending in the Philippines House of Representatives:
