Deputy Communications and Information Minister Nezar Patria addresses the media in a press conference after an artificial intelligence (AI) focus group discussion (FGD) on Nov. 27, 2023. - Photo from Jakarta Post/ANN
JAKARTA (Jakarta Post/ANN): As countries around the world rush to create regulations on artificial intelligence (AI) in light of a surge in real-life applications, Indonesia aims for its own, “local” approach.
However, experts have raised questions about the aspiration, considering the country’s predominant position as an AI consumer rather than a trailblazer in the industry.
Three years after introducing the National AI Strategy 2020-2045, the government has decided to issue an open letter to businesspeople as an initial step toward regulating AI tech.
A four-page draft of the letter shared by officials on Monday proposes ethical guidelines to shape company-internal policies for AI programming, analysis and consulting.
“The letter serves as an ethical guide that is not legally binding but rather governs at a normative level,” Deputy Communications and Information Minister Nezar Patria told the media in a press conference after an AI focus group discussion (FGD) on Monday.
The letter is to be finalized and released later this month.
Keeping things flexible
Despite its nonbinding character, the letter was not an attempt to come up with the feeblest possible measure, Nezar said.
Rather, he noted, given the rapid technological progress, the government had to remain open to any possible developments, which included being prepared to establish legally binding regulations as needed.
Nezar pointed out that Indonesia’s strategy diverged from that of the United States, the European Union and China, which were adept in AI technology, and would instead prioritize the local context in line with the country’s ideology.
“We can comply with global governance values while looking at our local values, such as cooperative inclusivity, humanity, safety, democracy, transparency, credibility and accountability,” he elaborated.
Indonesia has teamed up with AI Singapore to develop an open-source large language model (LLM) designed specifically for the Indonesian language and culture.
AI must understand Indonesia In a panel discussion following the partnership signing ceremony in Jakarta on Thursday, William Tjhi, head of applied research for foundation models at AI Singapore, highlighted the importance of this collaboration in developing AI solutions that resonated with South-East Asia’s diversity and prioritized cultural sensitivity as well as diverse use cases.
“We aim to align values on security and cultural sensitivity. Understanding what’s sensitive in another country requires local expertise, which is why having Indonesian partners is crucial,” William said.
Experts have commended the ethical guidance as an initial step toward regulating AI but also noted the need for a deeper understanding on that part of the government about the ecosystem and requirements for cohesive governance involving the public.
“There’s a good point in the letter, indicating that AI technology shouldn’t replace all existing manpower through automation,” Communication and Information System Security Research Center (CISSReC) chairman Pratama Persadha told The Jakarta Post on Tuesday.
On the other hand, as a guideline, it should address targets for system providers and users, especially for developers utilizing external models or application programming interfaces (APIs), he said.
Additionally, the guideline should highlight the benefits of AI, particularly for public interests. “The guidance shouldn't only be aimed at businesses but also include points for the general public, considering their increasing familiarity with this technology,” emphasized Pratama.
Rules needed, but which ones?
Indonesia Cyber Security Forum (ICSF) chairman Ardi Sutedja said on Tuesday that the government still had “a lot to explore” before it could release ethical guidelines on AI, as the country’s tech industry was dominated by consumers rather than developers.
“Regulations are important, but the question is, what kind of regulation do we actually need?” Ardi said, noting that the letter was a guideline rather than a directive. Dirgantara Adisutjipto Institute of Technology (ITDA) rector and AI expert Arwin Datumaya Wahyudi Sumari echoed a similar sentiment.
“We're a nation of consumers, and that’s also the perception the world has of us. It's time to shift our mindset from being consumers to becoming producers,” Arwin told the Post on Wednesday.
He advocated for the release of a more comprehensive, legally binding regulation to foster AI development within the country. Arwin and his team are behind Indonesia’s National AI Strategy 2020-2025.
Published in 2020, the strategy focuses on five topics: health care, bureaucratic reform, education and research, food security as well as mobility and smart cities. One of the conclusions drawn by the team when drafting the strategy was the need for a presidential regulation to regulate AI, according to Arwin.
“We need a presidential regulation, and now it’s still in the works by the National Research and Innovation Agency [BRIN]. That will establish a clearer and broader framework for AI,” Arwin said. Arwin collaborated with BRIN in 2021 to create a secure testing environment, also called a sandbox, for AI systems.
That facility, Arwin said, had helped advance health care and could be used by other sectors like defense and finance, but it still lacked coverage and support.
Arwin urged government support for the AI sandbox ecosystem, arguing that it brought together industry, academia and the community as it could address the uncertainties associated with new technologies without holding back innovation.
A local approach
Indonesia is not alone in pursuing a softer approach to AI regulation, as opposed to detailed rulemaking in some countries that focus on specific concerns regarding AI tech. The US released an executive order on Oct. 30 outlining principles for responsible AI development that emphasizes safety, security, reliability and trust.
As a decentralized approach, it aims to foster innovation while still addressing national cyber threat concerns.
AI regulations to be introduced in Japan by the end of the year are expected to be closer to the US approach, prioritizing economic growth and solidifying the country’s position in advanced chip and AI technology.
In contrast, the European Union's AI Act, which is in the final stages ahead of approval, represents a comprehensive top-down approach to AI regulation, prohibiting the use of AI systems deemed to pose unacceptable risks, such as autonomous driving, employment hiring and law enforcement.
China, issued temporary regulations on Aug. 15 mandating that algorithms undergo prior state scrutiny and must align with “core socialist values”.
The regulation addresses issues such as fake news and prevents companies from applying dynamic pricing. - The Jakarta Post/Asia News Network