Left to right: Tony Close (Director of Global Content Regulation, Meta); Min Aung, (Assessment and Accountability Manager, GNI); Chenie Yoon (APAC Content Regulation Lead, Google); Dr. Carol Soon (Principal Research Fellow – Institute of Policy Studies, National University of Singapore)
By Min Thu Aung, GNI Assessment & Accountability Manager
The Online Safety Forum (OSF) is an annual event held by the Asia Internet Coalition, which aims to gather key stakeholders from the Asia-Pacific (APAC) region to discuss regulatory frameworks, emerging trends, and technological innovations in online safety and scam and fraud prevention. The 2024 OSF was held in Singapore between 12-13 December 2024. GNI participated in a panel that examined evolving trends on intermediary liability, AI-driven moderation, over-the-top provider (OTTs) regulations, and age assurance policies, highlighting the challenges of balancing online safety with freedom of expression. GNI’s presentation addressed five current and emerging content regulation trends that have had or are likely to have negative impacts on human rights in APAC, as well as the role of GNI and its Principles on Freedom of Expression and Privacy (“GNI Principles”) amidst these challenges.
Overbroad Laws and Regulations
Countries in APAC are increasingly enacting overbroad social media regulations. For example, mandates for social media licensing (currently unique to the APAC region) have been enacted in Malaysia (as announced on 1st August 2024 and effective 1st January 2025), and Vietnam (as per Decree 147, effective 25th December 2024) with limited stakeholder consultation. Regulators and industry actors in jurisdictions such as Sri Lanka, India, and Bangladesh have previously discussed (but have not yet mandated) similar measures. Overbroad content regulations in APAC share common features, including but not limited to:
- the need for platforms to take down content vaguely identified as “indecent”, “obscene”, “toxic”, “false”, “menacing” or “hatred” (among others);
- short notice periods for takedowns (e.g. 3-hour takedowns for “illegal” live-streamed content in Vietnam’s Decree 27 and 72);
- financial penalties and criminal liabilities for company employees in case of delays or non-compliance on takedowns.
The features mentioned above are set within APAC country contexts which generally exhibit the following:
- varying level of protection for human rights,
- broad executive powers to politically appointed Ministers,
- limited accountability or judicial review of legislative or executive actions,
- varying levels of judicial independence,
- limited regulator independence, and
- a history of criminalization of online speech.
The combined impact of overbroad regulations within these country contexts is a high likelihood for over-moderation of content by tech companies and a chilling effect on the speech of citizens, both of which negatively impact freedom of expression. GNI has historically and regularly advocated against overbroad laws and regulations throughout its 16-year history and regularly assesses its company members’ responses to them in line with the GNI Principles through its assessment process.
Mandatory Use of AI in Content Moderation
To support ever shorter notice periods for takedowns, governments in the region are increasingly mandating AI-based content moderation. In India, the Intermediary Guidelines require major social media platforms to use AI tools to detect broadly defined content, including speech that purportedly threatens “public order”, “morality”, “sovereignty”, or “security”. Similarly, Vietnam’s 2022 regulations empower the Ministry of Public Security to block platforms that fail to remove “toxic” content within 24 hours, and they have explicitly demanded that companies adopt AI to meet these requirements. These actions lead not only to concerns of over-moderation but also bias inherent in the use of AI for such moderation, thereby negatively affecting both the right to freedom of expression and the right to non-discrimination. GNI assesses how company members conduct human rights due diligence on among other things the use of AI generally, including in the context of content moderation through its ongoing assessment process.
From Age Verification to Real Name Registration
Age verification is another regulatory trend not unique to APAC but with an especially elevated risk of negative impact on rights in this region due to the previously mentioned challenging country contexts.
Australia is planning to mandate age verification for social media and adult sites by November 2025. Less rights-respecting jurisdictions plan to enact similar measures, including neighboring Papua New Guinea. India’s Data Protection Act mandates parental consent for processing data of individuals aged under 18, thereby requiring data processors to verify the age of their data subjects.
These emerging age verification mandates resemble the trend of real name registration in APAC in that they stipulate the provision of personal identification or biometrics to access digital services. China and South Korea were the first two countries in the world to implement de jure real name registration (though such registration was rolled back in South Korea in stages). While the age verification regulations mentioned earlier amount to de-facto real-name registration, Vietnam’s Decree 147 restarts the trend initiated in East Asia by explicitly mandating real-name registration.
While such approaches are framed by governments as pursuing legitimate and necessary aims (such as child protection and cybercrime prevention), concerns have been raised about disproportionate government surveillance and data security risks in less rights-respecting jurisdictions that may impact the right to privacy, a chilling effect on freedom of expression due to the loss of online anonymity, and a risk of discrimination if AI is used for age verification or those with no identification documents.
In its assessment process, GNI assesses how company members address such requirements, including by:
- advocating against such policies,
- conducting human rights due diligence on the impact of such verification systems,
- implementing the least intrusive method of verification,
- providing transparency on age verification methods and algorithm audits (if AI is used for verification), and
- providing grievance mechanisms for users.
Government control of generative AI (genAI)
Another emerging content regulation trend in APAC (and globally) is government control of genAI through bans of specific platforms, licensing, and regulations on training data and outputs (among others). For example, China blocked ChatGPT in March 2023 due to its ability to circumvent censorship and has required local genAI services such as Tencent’s ErnieBot to hold licenses and “reflect core socialist values” that avoid politically sensitive topics. India followed suit in March 2024, mandating government approval for new AI tools before its 2024 general election. Government control of genAI has an unequivocal negative impact on the rights to freedom of expression, the right to non-discrimination, and potentially the right to privacy. GNI will publish recommendations on rights-respecting approaches by governments and companies on AI-related regulations and government orders in 2025 and continue to work with our members to understand how the GNI Principles can continue to guide multi stakeholder advocacy and responsible business conduct related to genAI services.
An even more challenging global context
The challenges identified above could be exacerbated in the near future, as geopolitical tensions and multilateral developments influence the way governments in the region approach tech regulation. For instance, the UN is poised to adopt a new Convention against Cybercrime (“Convention”). While attempting to address legitimate challenges related to the investigation and prosecution of cybercrimes, as GNI has pointed out, the Convention is also likely to be misused by governments who seek to consolidate state control over data, networks, and digital services, as well as those eager to perpetuate transnational repression under the guise of criminal prosecution.
The role of GNI
During the panel, GNI explained that addressing the challenging issues mentioned above in APAC requires collaborative, multistakeholder problem-solving and collective advocacy toward governments on protecting human rights and promoting accountability. GNI has promoted a holistic multistakeholder approach in its regional policy advocacy in Malaysia, Sri Lanka, and Bangladesh (among others), and in its contribution to initiatives such as the UN B-Tech GenAI project and the OECD.AI Network of Experts to advance respect for human rights both state and corporate actors.
On the panel, GNI also highlighted its approach to company accountability through its unique multi stakeholder assessment process that reviews its company members’ implementation of the GNI Principles every 2-3 years and emphasized how its current cycle of assessments will incorporate the company members’ responses to challenges such as those highlighted in this article.
GNI concluded with a call to action inviting actors in the region to join together in support of multistakeholder advocacy towards APAC governments for proportionate and balanced laws, regulations, and orders that address only legitimate and necessary aims, while simultaneously promoting company accountability and industry learning on rights-respecting responses to these laws, regulations, and orders.