As AI regulation looms, should Asia follow West's lead and risk 'killing innovation'?

British Prime Minister Rishi Sunak announced this month that Britain will host the world's first global summit on artificial intelligence safety later this year, with "like-minded" Asian countries and companies set to take part in discussions on regulating the rapidly developing technology before it is too late.

The move comes as lawmakers in Europe recently signed off on the world's first set of comprehensive rules for AI, adding to a growing chorus of calls for global regulation that shines a light on the efforts of Asia's tech hubs.

Earlier this month, OpenAI CEO Sam Altman visited India, which currently holds the rotating presidency of the G20, and urged the country's officials to help shape an international regulatory framework. His US-based firm developed the AI chatbot ChatGPT.

Do you have questions about the biggest topics and trends from around the world? Get the answers with SCMP Knowledge, our new platform of curated content with explainers, FAQs, analyses and infographics brought to you by our award-winning team.

But analysts say any laws regulating the technology must not also stunt its development, especially in Asia where many remote regions are underserved by public services at present and the innovative use of AI could transform lives.

"I believe that a wait-and-watch approach may be the most practical at the moment," said Amit Joshi, a professor of AI, analytics and marketing strategy at the international Institute for Management Development, based in Singapore and Switzerland.

"Hubs like Singapore need to put in place basic guidelines and guardrails around the misuse of these technologies," he said. "However, without really understanding these technologies' future trajectory, there is a risk of overregulation or misplaced rules killing innovation."

The city state's Personal Data Protection Commission, established in 2013, currently regulates information privacy and the use of AI in Singapore, including overseeing developers and other companies.

Lee Wan Sie, director of artificial intelligence at Singapore's Infocomm Media Development Authority - an autonomous government agency under the Ministry of Communications and Information - told CNBC on June 19 that the city state was not looking to regulate AI.

"At this stage, it is quite clear that we want to be able to learn from the industry. We will learn how AI is being used before we decide if more needs to be done from a regulatory front," she said, adding that regulation may be introduced at a later stage.

The clamour for global AI regulation, however, is bound to put pressure on Asian nations to get on board sooner rather than later, analysts say. Britain's coming summit is expected to discuss how risks can be mitigated through "internationally coordinated action".

Duplicating the regulations of an advanced economy such as Britain or those of a varied region like the European Union would be much easier to do, but "this may equally freeze out" companies from other nations, such as China or India, AI professor Joshi said.

Copying other regions' regulations would also make it harder for Asian nations to customise laws to local needs, he said, which may hinder domestic companies' development.

China is expected to have a draft of its AI regulations ready for review by the country's lawmakers within the year. But it remains to be seen whether the world's second-largest economy will participate in this year's UK summit amid simmering tensions over the sharing of technology.

It's also unlikely that India will rush to join an international framework any time soon, although industry body the National Association of Software and Service Companies (NASSCOM) did release guidelines this month for the use of AI by developers and researchers, according to insiders.

The South Asian nation has no specific laws for AI but the government is poised to release a draft of its Digital India Act, which is intended to replace the old Information Technology Act that came into play in 2000.

It's not yet known what the proposed law could mean for AI, but IT Minister Rajeev Chandrasekhar said earlier this month that any regulation will be to protect against "user harm".

"It is likely that in the near future, India will regulate AI with a light-touch approach via existing laws, rather than going for AI-specific regulation," said Siddharth Mahajan, a partner at Athena Legal, a New Delhi law firm.

Sanjay Kaushik, managing director of security and risk management consultancy Netrika Consulting India Private Ltd, said that Prime Minister Narendra Modi's government was unlikely to be swayed by global calls for regulation as a separate law would be "jumping the gun" on a technology that is still evolving.

According to a NASSCOM report this month, the adoption of AI could add US$500 billion to India's economy by 2025.

"India has attracted companies and positioned itself as a potential service hub for AI," said Soumen Datta, a business analytics partner at BDO Digital, which consults on artificial intelligence. However, with the global trend towards AI regulation, India "will also need to adapt its approach in the future", he said.

Most Southeast Asian nations currently have in place a mix of voluntary codes of practice and industry-specific rules, regional lawyers told This Week in Asia.

Lauren Hurcombe, a Hong Kong-based partner specialising in technology at global law firm DLA Piper, said it was "increasingly clear that light touch regulation is unlikely to be sufficient" given the degree of misuse of AI systems, particularly generative AI, which can produce text, pictures and other media in response to prompts.

One of the biggest concerns surrounding AI are so-called deepfakes, which feature audio or visual content that's been crafted or manipulated - sometimes to run disinformation campaigns that can quickly go viral online.

"We anticipate that governments in the region will begin taking steps towards regulating AI in general in the next few years," said Hurcombe, citing the example of Thailand, which is working on a draft regulation. The EU's looming AI laws have set the ball rolling, she said.

Others say that delays were always going to be inevitable in the process of drafting and coordinating global regulations.

John Collins, head of AI for Asia at FTI Consulting, gave the example of banking regulations, which took decades to become harmonised international rules after the first Basel Accords were drawn up in Switzerland in 1988, and are still evolving.

"There is no one-size-fits-all approach that can be considered universally optimal. Put another way, what is right for the UK is not necessarily right for Hong Kong and Singapore."

But he added that appropriate regulations were generally considered necessary.

Governments need to be aware of the potential legal issues around the use of AI and look at potential solutions, another lawyer said.

"Generative AI is only as good as the data it learns from, and the use of such AI raises important questions as to whether such learning and the output of the AI potentially infringes the intellectual property of others," said Nicholas Lauw, a Singapore-based partner in technology and digital assets at RPC Premier Law.

"At the same time, people need to be aware that the use of AI may expose them and their employers to legal risks."

While Britain is now talking about setting up "strong guardrails" for AI, a UK White Paper on the technology published on March 29 suggested a relatively light touch approach, Lauw said, noting that it would also be best to see what comes out of Britain's summit to be sure of the direction of the proposed regulatory framework.

"Given that Rishi Sunak is seeking to find a 'global' solution, he might produce a moderate solution worth adopting," he added.

This article originally appeared on the South China Morning Post (SCMP).

Copyright (c) 2023. South China Morning Post Publishers Ltd. All rights reserved.