|

Asia Pacific Regional Internet Governance Forum 2026 Synthesis Document – Draft 1

The Process 

1 Leave a comment on paragraph 1 0 The Asia Pacific Regional Internet Governance Forum (APrIGF) 2026 Synthesis Document  brings together key discussions, ideas, and outcomes emerging from the APrIGF meet, held during October 4-7, 2026, in Kathmandu, Nepal, in hybrid mode. With the overarching Forum theme of “Safeguarding a Resilient Internet in Times of Crisis”, it  issues that matter to the region and contributes to wider discussions on the future of Internet governance. APrIGF 2026 took place against the backdrop of recent natural calamities  in Nepal, causing widespread loss of life, displacement, and infrastructure damage. 

2 Leave a comment on paragraph 2 0 Over the years, the Synthesis Document has become an important knowledge resource of APrIGF and has encouraged similar documentation efforts by other national and regional Internet governance initiatives.

3 Leave a comment on paragraph 3 0 The document is being prepared by the 2026 Drafting Committee, with support from the APrIGF Secretariat. Contributions from the wider APrIGF community are  invited through two public input rounds and the APrIGF 2026 Townhall sessions. Participants have shared  their comments and suggestions through the APrIGF online consultation platform https://comment.aprigf.asia during the input periods, helping to strengthen and reflect the diverse regional perspectives in the final document.

Preamble

4 Leave a comment on paragraph 4 0 The APrIGF 2026 has brought  together stakeholders from the Asia-Pacific region to discuss key issues that are shaping the future of the Internet and digital governance. The Forum was held in Nepal at a time when Internet resilience is being challenged on multiple fronts, including during the recent natural catastrophe in Nepal and the efficacy of subsequent crisis response mechanisms.

5 Leave a comment on paragraph 5 0 APrIGF 2026 focused on the overarching theme: “Safeguarding a Resilient Internet in Times of Crisis.” The theme recognised the need to keep the Internet available, reliable, secure, trusted, and inclusive during crises such as natural disasters, conflicts, cyber threats, and infrastructure disruptions.

6 Leave a comment on paragraph 6 0 The 2026 discussions focussed on four key sub-themes:

  1. 7 Leave a comment on paragraph 7 0
  2. Future of APrIGF & Multistakeholder Ecosystem
  3. AI Governance for Good
  4. Digital Resilience and Building of Trust
  5. Universal and Meaningful Connectivity

8 Leave a comment on paragraph 8 0 APrIGF 2026 has provided an equitable space for governments, civil society, the technical community, the private sector, academia, youth, and other stakeholders to share experiences and collectively contribute to a resilient and inclusive digital future for the Asia-Pacific region.

9 Leave a comment on paragraph 9 0 Relevant inputs from other sessions and TownHalls are incorporated as the Synthesis Document develops.

10 Leave a comment on paragraph 10 0 Track 1: Future of APrIGF & Multistakeholder Ecosystem

11 Leave a comment on paragraph 11 0 The permanent mandate of the global Internet Governance Forum (IGF), confirmed through the 20-year review of the World Summit on the Information Society (WSIS+20 Review) in 2025, provides greater continuity for Internet governance processes. For APrIGF, it raises questions about how to sustain participation beyond the annual Forum, connect national, sub-regional and youth initiatives, and bring Asia-Pacific priorities into global discussions.

12 Leave a comment on paragraph 12 0 Track 1 discussions explored year-round engagement, links between local and global governance processes, and multistakeholder cooperation in crisis preparedness. The recent floods in Nepal provided a case study of Internet resilience and the importance of coordination before crises arise.

  1. 13 Leave a comment on paragraph 13 0
  2. The Role of APrIGF under the Permanent IGF Mandate

14 Leave a comment on paragraph 14 0 1.1 A year-round regional platform

15 Leave a comment on paragraph 15 0 The IGF’s permanent mandate under the United Nations (UN) aegis provides institutional continuity and allows longer-term planning at global, regional, and local levels. Equally, this provides a wider opportunity for APrIGF to extend engagement beyond its annual meeting by connecting national, sub-regional, and youth initiatives throughout the year. Discussions and reports are and could contribute to ongoing processes, including the Global Dialogue on AI, and  national and regional IGFs providing  support for the implementation of World Summit on the Information Society (WSIS) action lines.

16 Leave a comment on paragraph 16 0 Criticism that IGFs do not produce negotiated outcomes was acknowledged. Its value however, lies in bringing governments, industry, the technical community, civil society, and youth together on an equal footing. Such exchanges allow stakeholders to understand different perspectives and inform policymaking. The challenge for APrIGF is to sustain these exchanges throughout the year and connect their outcomes to ongoing policy processes, regionally and globally.  

17 Leave a comment on paragraph 17 0 1.2 Linking Local, Regional and Global Processes

18 Leave a comment on paragraph 18 0 APrIGF was described as a link between local initiatives, regional discussions and global Internet governance processes. National, local, and youth initiatives can bring different approaches to the shared challenges into regional discussions. Comparing these experiences can help identify common concerns and useful practices, while also recognising differences in local needs and conditions. This can help identify shared concerns and useful practices without assuming that one solution fits the entire region. APrIGF can carry these diverse regional perspectives into global discussions while sharing developments from those processes with national and local initiatives.

19 Leave a comment on paragraph 19 0 The Synthesis Document is one potential channel for sharing regional findings, with a proposal to build a repository of best practices for policymakers and other stakeholders. Other proposals include focusing on one or two priority issues annually, discussing them through local hubs, and establishing working groups to prepare papers and exchange experiences. These efforts could build on the regional working group established for the 2025 WSIS+20 Review and provide opportunities for continued engagement beyond the annual forum.

20 Leave a comment on paragraph 20 0 1.3 Resilience in times of crisis

21 Leave a comment on paragraph 21 0 The recent floods in Nepal highlighted broader challenges for Internet resilience across the region. Reliable electricity, alternative connectivity infrastructure and community networks remain crucial, particularly in rural and remote areas. Concerns about affordability and misinformation are also top of mind in the region. Improved digital and media literacy and stronger infrastructure resilience standards are required to address these concerns.

22 Leave a comment on paragraph 22 1 Experiences from disaster response highlighted the value of cooperation among governments, technical experts, civil society and local communities. Greater involvement of Internet service providers in preparedness discussions and better use of local expertise would both improve local disaster response and recovery efforts. The wider IGF community could also explore the adoption of a shared accountability framework for times of crisis.

23 Leave a comment on paragraph 23 0 Local actors remain the first responders. APrIGF can support preparedness by bringing stakeholders into discussions before crises occur, facilitating exchanges on resilience practices across the region, and connecting regional lessons to global IGF processes. 

24 Leave a comment on paragraph 24 0 Track 2: AI Governance for Good

25 Leave a comment on paragraph 25 1 Artificial Intelligence (AI) is an important emerging tool, and has the potential to act as a key enabler for sustainable development, productivity, and technological advancement. However, ensuring AI technologies serve the public good remains a critical challenge. Even the definition of  “AI Governance for good” will differ between countries and regions. The Asia-Pacific (APAC) region envisions an advanced and responsible model of AI Governance that  addresses broader societal, economic, cultural, and ethical implications, particularly with regard to algorithmic bias and unequal access to AI opportunities. Building public trust by embedding human rights into the development of AI governance through transparency and accountability is a central focus. Consequently, the use of AI in the sensitive, critical tasks of crisis response and peace-building has been met with scrutiny. Robust data governance and regulatory readiness to support information integrity and inclusive innovation are also priorities. The AI digital divide must be addressed by improving digital skills and capacity, without compromising on the importance of safe, secure, and trustworthy systems.

26 Leave a comment on paragraph 26 0 1. Principles for Responsible AI Governance

27 Leave a comment on paragraph 27 0 1.1 AI Bias

28 Leave a comment on paragraph 28 0 Algorithmic bias in AI has been shown to emerge through nonrepresentative training data, system design, model outputs, and decisions on how AI is deployed. But AI bias is not a purely technical problem; it reflects the many inequities perpetuated by humans throughout history and into the modern day. AI systems absorb and amplify  bias, social inequality, stereotypes, and historical prejudices that are present, even subtly, within their training data. As a result, AI can repeat or even strengthen existing biases related to gender, age, language, culture, ethnicity and social position. For example, in depicting a South Asian woman in leisure, AI might show a woman who’s cooking instead of traveling. This output shows narrow gender and cultural stereotypes and illustrates the impact of algorithmic bias on how communities are represented, perceived, and treated in digital environments. AI bias can create more concrete harms as well – AI tools trained on historical employment data may favour men and disadvantage women and younger candidates, while in healthcare, the underrepresentation of local languages and communities in medical datasets can result in inaccurate recommendations, cultural misunderstandings, and privacy risks. As AI adoption grows globally, these risks must be addressed and mitigated to prevent substantial, wide-scale harms.

29 Leave a comment on paragraph 29 0 AI governance in the Asia Pacific region should recognize that bias  comes not only from technology but also from society. Rules therefore should not focus only on algorithms, bias screenings, or technical standards, though these factors remain important. On a broader scale, it is crucial to address the root causes that sit beneath the issue of bias in AI systems Issues of human rights, social inequality, language, culture, access to technology, and participation in decision making must be considered when seeking to eliminate harmful AI bias in practice. 

30 Leave a comment on paragraph 30 0 1.2 High-Risk Usage & Human Oversight

31 Leave a comment on paragraph 31 0 High risk, high impact applications of AI must be addressed through comprehensive and binding international governance. 

32 Leave a comment on paragraph 32 0 The absence of enforceable international rules relating to AI usage in military contexts raises serious concerns about meaningful human control, accountability, and human rights. AI implementation accelerates the risks of technology-facilitated psychological distance in military decision making, potentially resulting in reduced human awareness and oversight of consequences. This increased use of AI in military and conflict settings raises a critical question on who should be held responsible, especially if the decision of the AI system results in someone’s death. The reliability of digital evidence during conflicts also raises concerns as AI-generated military footage and manipulated satellite images can influence public opinion, undermine trust, and complicate the use of digital information in international discussions.

33 Leave a comment on paragraph 33 0 Existing laws in this area fail to address this gap, such as the EU AI Act which does not apply to military or national security systems. Asia-Pacific is particularly vulnerable to the governance gap, with rapid development and adoption of AI often outpacing the speed of lawmaking and regulatory oversight. Regional tensions and active conflicts within Asia-Pacific also increase the urgency of this issue, particularly for individuals in the affected regions. An international and binding framework for responsible use of AI in conflict and military contexts must be developed to fill this gap.

34 Leave a comment on paragraph 34 0 Beyond military applications, human oversight must be maintained throughout the AI lifecycle. Regular ethical audits, transparency, and accountability mechanisms to identify and mitigate bias and discriminatory outcomes must be integrated, particularly in high-impact sectors such as healthcare and recruitment. But these mechanisms alone may not be enough in all contexts; for example, in healthcare and some other public services, decisions should not depend only on an automated system. AI may assist in processing information or identifying patterns, however, the final decisions must remain subject to human judgement. Effective AI governance of these crucial, high-risk areas must be binding,  consider meaningful human oversight, and result from inclusive international cooperation. Technological progress should always  remain accountable to human rights and international humanitarian law.

35 Leave a comment on paragraph 35 0 1.3 Inclusivity and Regional Relevance

36 Leave a comment on paragraph 36 0 Most AI systems are still built and controlled outside the Asia-Pacific or designed outside regional frameworks or guidelines, using data and standards that come from institutions and are developed with minimal consultation of diverse groups. Relevant authorities must establish policies that ensure the involvement of all stakeholders from the initial design and development stage. This includes youth, civil society, developing economies, women, people with disabilities, and the many other groups whose lives are heavily impacted by the direction of AI development.

37 Leave a comment on paragraph 37 0 Meaningful participation from diverse communities in shaping how AI is designed, regulated, and deployed is the most important element of  inclusive AI governance. Rather than being consulted only after policy decisions have been made, underrepresented and diverse groups should be given the opportunity to contribute their unique experiences and voices throughout the AI lifecycle. Asia Pacific is a highly linguistically and culturally diverse region. This increases both the resources required to develop locally relevant AI models, but also the importance of doing so. It is crucial to create inclusive datasets, strengthen local language resources, and support sovereign AI models that reflect local priorities, cultures, and contexts. The input of small and developing economies must also be considered with equal weight, to ensure that international AI governance frameworks are inclusive and responsive to their needs.

38 Leave a comment on paragraph 38 0 1.4 International Alignment

39 Leave a comment on paragraph 39 0 Asia-Pacific is a diverse region with vast, often differing priorities when it comes to AI governance. Each economy is focused on addressing its own issues of national significance, leading to fragmented approaches and a complex regulatory market for international operators. This makes enforcement difficult, and limits the influence of any one approach due to the many competing frameworks of various scope. Aligning on regional priorities for Asia-Pacific and developing a shared set of rules is a key step to tackle this issue. A unified regional stance will give Asia-Pacific strength in further international discussions and help ensure that local interests are reflected globally. It will also progress the creation of accountable and meaningful AI governance which has the ability to impart real change.

40 Leave a comment on paragraph 40 0 International cooperation on AI also extends to data governance and privacy. Digital services do not stop at regional borders, and AI training data is often sourced from a range of countries without regard for their national legislation or requirements for data use. A shared approach to data use, sharing, and privacy is required to ensure that end users are not disadvantaged by opaque usage of their personal data by global digital actors. This could include developing methods to track data use across borders, adoption of shared principles, definitions, and standards on personal and community data rights, and regional cooperation to develop local models that respect data privacy and intellectual property rights.

41 Leave a comment on paragraph 41 0 While it is important for countries to adapt international standards to their local contexts, it is also crucial to strengthen regional cooperation through shared principles, knowledge exchange, and interoperable standards.  When regional partners with similar interests work together, their voices can be amplified beyond the influence of any one economy alone. Strategies to further this vision include complementary or in-kind capacity building support, providing funding and technical expertise, sharing resources, and ultimately seeking to advance meaningful participation of Asia-Pacific in regional and international decision making.

42 Leave a comment on paragraph 42 0 2. AI and Information Integrity 

43 Leave a comment on paragraph 43 0 2.1 Transparency 

44 Leave a comment on paragraph 44 0 AI systems depend on ingesting vast amounts of data, yet lack transparency around data collection, use, and protection.. Much of the data used to develop and train AI systems is personal in nature, making privacy and data protection a core part of responsible AI governance. This lack of clarity can also weaken public trust in AI systems and the companies that create them., It is important for AI creators to be transparent about data practices, obtain meaningful consent, limit data use to clear and legitimate purposes, and keep information secure throughout the AI lifecycle. Countries within Asia-Pacific need practical, context-sensitive frameworks that protect data rights while still supporting inclusive innovation. It is important to also have a clear accountability to seek redress when someone’s data is misused or their privacy is violated. For example, it should be possible to track how AI is built, including through searchable records of training data and the ability to delete and have the AI un-learn specific information when requested. Users should be able to clearly understand where, how, and when their data is being used, and be given the option to opt out at any time. With the increased value and commodification of personal data in an AI-driven world, it is more important than ever to ensure it is handled with transparency and respect for individual privacy.

45 Leave a comment on paragraph 45 0 Transparency requirements should not just apply to the AI training process, but also to its implementation and impact. As AI assistants increasingly stand between users and the wider Internet, a few platforms risk becoming the whole Internet for many new users. This raises questions about who decides what people see and whether users can move between services or challenge an AI-generated answer. A lack of transparency surrounding the decision-making process of both AI systems and the companies that create them may have a concerning impact on the free Internet. An open and transparent Internet in the age of AI is one that preserves choice and makes available clear information on how different AI systems work, so that AI becomes another layer of opportunity rather than a gatekeeper or single source of truth.

46 Leave a comment on paragraph 46 0 2.2 Electoral manipulation and deepfakes

47 Leave a comment on paragraph 47 0  AI image and video generators have become increasingly convincing, reaching a point where ordinary people are often unable to distinguish between genuine and AI-generated materials. The many risks that come with this technology can be examined through the lens of the emerging challenge of AI governance for electoral integrity.

48 Leave a comment on paragraph 48 0 AI-generated misinformation, including deepfakes, have the potential to impact electoral integrity in the Asia-Pacific region These challenges are compounded by gaps in election laws, limited institutional capacity, unequal digital literacy, weak cybersecurity, and insufficient transparency around political advertising, campaign-related AI use, and platform content moderation. Different stakeholders have different roles to play in order to adapt to this environment. Platforms should provide clearer explanations for content removal, effective appeal mechanisms, and greater accountability for moderation decisions. Election authorities and governments might need stronger technical expertise, resources, and procedures to detect and respond to AI-driven manipulation. 

49 Leave a comment on paragraph 49 0 This should be elevated by improved cybersecurity for government systems and electoral infrastructure, inclusive voter education for vulnerable groups,and stronger cooperation among stakeholders on emerging methods to address these risks.. Building digital literacy and providing practical training for citizens, institutions, and communities to identify AI-generated content is also an important tool for reducing the impact to electoral integrity. The necessity to adopt updated AI-specific electoral regulations covering deepfakes, political advertising, disclosure requirements, and electoral manipulation has grown in light of recent technological advances. Human rights, transparency, oversight, and accountability should underpin the use of AI in elections, with the protection of electoral integrity as a shared objective.

50 Leave a comment on paragraph 50 0 3. Access, skills, and capacity

51 Leave a comment on paragraph 51 0 3.1 Digital Accessibility and Inclusive AI
The mere integration of AI into digital platforms does not guarantee universal accessibility. People with disabilities, who may require assistive technologies or rely upon accessibility standards to be able to access technology, are often overlooked by AI developers. Accessibility is often treated as an optional feature, or something that can be added in a future update, rather than a core requirement of use for the many people who rely on assistive solutions. Disabled voices are often underrepresented or absent in the companies that build and release these technologies, and training data may not proportionally reflect the experiences or realities of living with a disability. This can result in many AI tools being functionally unusable by this group, as it may be incompatible with their needs or simply not designed with their lived experiences in mind. 

52 Leave a comment on paragraph 52 0 Cost is another significant factor, and AI projects that actually seek to improve or act as assistive technologies rarely get enough funding to break through. Companies may also make the decision to not pursue or fund accessibility of their AI products or integration due to a perceived lack of profitability. Those who require accessibility may be seen as a comparatively small market whose highly specific needs are not necessary to accommodate from a profit standpoint.

53 Leave a comment on paragraph 53 0 A people-centric and inclusive approach to AI design and deployment is required to progress the goal of universal accessibility to technology. Practical solutions include government development of policy, laws and regulations to make accessibility a must for all digital and AI systems. Governments could even require certain accessibility standards to be fulfilled by their own contracted parties or projects. Other stakeholders, such as universities, NGOs, and private companies, should do their part by ensuring that their own materials and integration of AI are fully accessible, by design. Above all else, it is crucial to create comprehensive accessibility and inclusivity standards that everyone follows, so there’s consistency wherever AI comes into digital solutions.

54 Leave a comment on paragraph 54 0 3.2 AI Literacy

55 Leave a comment on paragraph 55 0 With the increased prevalence of AI in society, AI literacy is a basic skill and a necessity for all people. AI literacy goes beyond frequency of AI use – simply having access to AI technologies is not enough to ensure an individual or community can extract value from them, or conversely, be protected from the risks they pose. Individuals must understand the importance of knowing where AI-generated information comes from, whether it can be verified, whether it may be biased or manipulated, and how to identify deepfakes in particular. This applies whether they personally use AI for information, or if they are simply exposed to it on the Internet, something that applies to everyone.

56 Leave a comment on paragraph 56 0 There is a growing AI literacy gap, both when it comes to use of AI and distinguishing between authentic and AI-generated content. The gaps are widest in rural areas, among older adults, women, people with disabilities, and members of small language groups with little native-language support online. Even where physical devices and stable network connections are available, vulnerable communities may self-exclude from AI usage due to a lack of confidence or knowledge. Without development of AI knowledge and training for these groups, the digital divide may grow stronger and leave more people behind. However, finding a way to reach these communities and provide the appropriate resources can prove difficult. There is a lack of sufficiently skilled teachers and leaders to support the development of AI literacy skills, particularly within rural communities. Many existing learning materials on AI literacy and deepfake identification are also not available in local or indigenous languages, further increasing the difficulty for small or marginalized cultural and language groups to improve their own skills in this area. 

57 Leave a comment on paragraph 57 0 Trust in AI cannot rest on individual users alone, or rely on self-education to become confident and safe. AI literacy education is a key factor in solidifying user trust and preventing a widening digital divide. Furthering AI skills development particularly for at-risk groups is the responsibility of all stakeholders, including governments, schools, workplaces, and community organisations, each of whom have a unique role to play.

58 Leave a comment on paragraph 58 0 One suggested approach to address this issue is the development of public awareness campaigns that teach people how to determine whether content is AI generated, verify it using external sources, and share that knowledge with their local communities. Teachers and parents should also be provided with AI education, which they can then use to guide children through safe and effective use of AI tools. Physical learning spaces such as community learning hubs, libraries, and local fellowship programs can also help close the AI knowledge gap. Economies within the Asia-Pacific region must prioritise education and upskilling to ensure that their populations are not disadvantaged by the direct and indirect impacts of AI that affect them.

  1. 59 Leave a comment on paragraph 59 0
  2. Ethical alignment

60 Leave a comment on paragraph 60 0 4.1 Sustainability and Environmental Responsibility

61 Leave a comment on paragraph 61 0 With the rapid expansion of AI over the past few years, the resources required to support AI systems have scaled exponentially. From water and power usage, to the increased scale of hardware required to train and operate AI models, the environmental impact of such systems is significant. Their sudden growth has created increased demands on existing digital infrastructure and necessitated the accelerated development of additional infrastructure to handle the load.  AI expansion should not slow the adoption of sustainable and climate-friendly infrastructure and energy sources, and its environmental impacts must be properly considered and mitigated.Digital growth should help communities and boost economies without negatively impacting the environment or draining natural resources.

62 Leave a comment on paragraph 62 0 Governments and technology companies must prioritise long-term environmental sustainability over short-term gains associated with rapid AI development, and ensure full compliance with environmental standards and regulations. Digital infrastructure should work, be affordable, and be accessible, without compromising on climate impact. Energy efficiency of infrastructure, e-waste management of Internet devices, and investment in renewable energy are some of the strategies that various stakeholders can implement to help ensure that AI-related growth is sustainable. Governments in particular should back renewable energy, and push for innovation in the field of sustainability to support a future where the Internet can run cleanly, sustainably, and without harming the planet. 

63 Leave a comment on paragraph 63 0 4.2 Ethical and Human-Centric Design

64 Leave a comment on paragraph 64 0 Human-centric design places people, their needs, and their rights at the centre of the development and deployment of AI systems and digital technologies. AI must be designed to not only support human dignity, privacy, and equality, but also remain aligned with ethical frameworks and best practices. The needs of the real humans impacted by AI must also be considered at every step, from design all the way to rollout of digital platforms. The human-centric design approach may help address other key issues such as bias, exclusion, and automated decision making that lacks oversight. Even good ideas may be executed poorly without appropriate consideration of real human experiences at the center of the design process.

65 Leave a comment on paragraph 65 0 Human rights must be respected when designing and releasing AI systems. AI should never infringe upon people’s fundamental rights, and there must be appropriate checks and balances during the design process to prevent it from doing so. Civil society, policymakers, governments, and the AI companies themselves should collaborate to address these challenges and ensure that human rights are not eroded by the impacts of rapid growth of AI technologies. Human-centric design particularly requires the organisations developing these technologies to take responsibility for their real-world consequences, ensuring that innovation remains accountable to the people it is intended to serve. Organisations should encourage the creation of independent AI ethics committees and open feedback systems to support an ethical future for AI.

66 Leave a comment on paragraph 66 0 Track 3: Digital Resilience and Building of Trust

67 Leave a comment on paragraph 67 0 Digital and Internet resilience is not only about keeping networks operational in normal times but also during crises like natural disasters and conflicts . In both circumstances, these depend on whether people can access true, factual and reliable information timely and as when they need; use secure digital services and exercise their rights without fear of surveillance or exclusion, supported by institutions that are accountable and capable of responding to disruption. In the  Asia-Pacific region, natural disasters, energy shortages, cyber threats and government-imposed restrictions expose weaknesses in both digital infrastructure and its governance and raises resilience and trust questions. These risks are particularly significant for underserved communities and smaller economies with limited resources, digital capacities, and alternative means of connectivity.

68 Leave a comment on paragraph 68 0 This track examines five interconnected dimensions of resilience. First, it considers the risks to essential communication posed by infrastructure failures, energy shortages and gaps in local preparedness. Second, it examines the challenges of measuring Internet disruptions and making information on infrastructure vulnerabilities publicly accessible, alongside concerns about the environmental sustainability of digital expansion. Third, it addresses cybersecurity risks and the unequal impact of threats and security measures on different communities, including gendered harms. Fourth, it explores how growing reliance on digital public infrastructure (DPI) raises concerns about personal data protection, surveillance and public accountability. Finally, it examines how online harms and Internet shutdowns threaten fundamental rights, particularly for marginalised communities.

69 Leave a comment on paragraph 69 0 Together, these issues point to a common principle: Digital and Internet resilience should be assessed not only by whether digital systems remain operational, but also by whether people can access, safely use and trust the Internet, particularly those most vulnerable to disruption. Achieving this requires more than investment in technology. It depends on effective institutional safeguards, protection of fundamental rights and meaningful community participation with capacities in digital governance.

  1. 70 Leave a comment on paragraph 70 0
  2. Resilient infrastructure in conditions of disruption

    1.1 Infrastructure diversity and continuity of essential services

71 Leave a comment on paragraph 71 0 Resilient infrastructure must sustain essential communication when normal services are disrupted. During disasters or prolonged outages, the failure of a single network can isolate communities when access to warnings and assistance is most urgent. Maintaining connectivity therefore requires alternative communication pathways that can withstand different kinds of disruption.

72 Leave a comment on paragraph 72 0 Satellite connectivity, as a viable alternative, can extend access beyond terrestrial mobile coverage, while community radio can reach people without Internet access. These approaches serve different needs and can complement one another within a broader emergency communication system. A recent offshore pilot in coastal Bangladesh demonstrated the potential of satellite connectivity beyond the mobile coverage available during field testing.  However, a successful connection under test conditions does not establish continuous coverage or reliability during severe cyclones. Infrastructure planning, design and adoption, must account for these limitations and test how alternative systems perform under disruptions.

73 Leave a comment on paragraph 73 0 Backup arrangements must also consider their energy requirements. An alternative connection offers little protection if it cannot be powered. Lower-bandwidth systems that consume less electricity may preserve essential communication during extended shortages. Maintaining a basic service can be more useful than attempting to sustain full capacity until all connectivity is lost.

74 Leave a comment on paragraph 74 0 1.2 Integrating energy resilience into network design

75 Leave a comment on paragraph 75 0 Reliable power should be incorporated into network design from the outset. Equipment demand and local operating conditions determine the backup capacity required. Unstable electricity also presents risks: brownouts can damage equipment even when the supply has not failed completely. Protective measures and regular monitoring are therefore important alongside backup generation.

76 Leave a comment on paragraph 76 0 Decentralized renewable energy can support connectivity where grid electricity is unreliable. Solar-powered systems and local energy sharing offer opportunities to reduce dependence on fuel-based backup, but their effectiveness depends on appropriate engineering and sustained maintenance. They should be selected according to community needs rather than treated as universally suitable solutions.

77 Leave a comment on paragraph 77 0 Energy and telecommunications planning must consequently be coordinated. Integrating digital infrastructure into disaster-preparedness and climate-adaptation strategies can help address dependencies before emergencies occur. Public comments reinforce  the need to assess this resilience through the ability to maintain or restore meaningful communication during crises, rather than network coverage under normal conditions alone.

78 Leave a comment on paragraph 78 0 1.3 Affordable access and sustainable financing

79 Leave a comment on paragraph 79 0 Internet infrastructure cannot provide meaningful protection if people cannot afford to use it. Offshore satellite services demonstrated technical potential, yet recurring costs remained a significant barrier for example in case of the small-scale fishing communities. Elsewhere, people lost access to online information because they could no longer afford mobile data, despite functioning telecommunications networks.

80 Leave a comment on paragraph 80 0 Financing must therefore address continuing access as well as initial deployment. Subsidies and shared subscriptions were proposed to distribute costs more fairly. Community-based arrangements may make services accessible to people who cannot sustain individual subscriptions, provided responsibilities for operation and maintenance are clear.

81 Leave a comment on paragraph 81 0 Short-term project funding alone is insufficient to secure continuity. Longer-term public support and partnerships across the multistakeholder community could help sustain services in commercially underserved areas. Proposals to extend universal-service-style funding to community broadcasting similarly recognize that resilience requires investment in the information channels people rely on, alongside network infrastructure.

82 Leave a comment on paragraph 82 0 1.4 Trusted and accessible community communication

83 Leave a comment on paragraph 83 0 The practical value of infrastructure lies in the information it enables people to receive and exchange. During emergencies, warnings must be understandable and relevant to local circumstances. Familiar languages and established relationships can help community broadcasters connect official guidance with the decisions people need to make.

84 Leave a comment on paragraph 84 0 Trust requires sustained verification and accountability. Broadcasters have described consulting knowledgeable sources before sharing advice and engaging directly with listeners about misinformation. These practices depend on relationships developed through everyday work, rather than only after a crisis begins. Two-way communication also allows communities to report unmet needs and bring local concerns to decision-makers.

85 Leave a comment on paragraph 85 0 Digital platforms can extend the reach of community communication, while terrestrial broadcasting preserves an alternative for people without Internet access. Online streaming, however, cannot provide continuity during complete disconnection. Radio systems also require protection and backup power. A diverse communication environment must therefore account for the vulnerabilities of each channel and people’s ability to access it.

86 Leave a comment on paragraph 86 0 1.5 Local capacity, coordination and preparedness

87 Leave a comment on paragraph 87 0 Community skills are essential to making resilient infrastructure usable. Training local responders can help people operate equipment and share emergency information. Preparedness should also ensure that residents know which communication channels remain available when normal services fail.

88 Leave a comment on paragraph 88 0 Continuity arrangement facilities  need regular testing. Written plans alone do not establish whether equipment will function, responsibilities are understood, or communities can use the alternatives provided. Questions about locally powered Wi-Fi and data caching further identifies the possibilities for maintaining information exchange during complete disconnection, although these capabilities were not demonstrated by the offshore pilot.

89 Leave a comment on paragraph 89 0 Effective preparation requires sustained cooperation across the multistakeholder community. Community operators and broadcasters should participate in planning so that technical decisions reflect local conditions. Clear coordination and appropriate emergency licensing arrangements can support timely action. By connecting these measures with sustainable financing and trusted communication, resilience planning can better protect communities before, during, and after disruptions.

  1. 90 Leave a comment on paragraph 90 0
  2. Measuring Internet resilience

91 Leave a comment on paragraph 91 0 2.1  Measurement and transparency

92 Leave a comment on paragraph 92 0 Building Internet resilience requires reliable evidence of how networks perform during disruption. Infrastructure may appear operational while users experience service failures, degraded connections or restrictions on access. Open measurement platforms using community-hosted probes can help identify these problems by monitoring network reachability and changes in traffic routing.

93 Leave a comment on paragraph 93 0 However, measurement coverage remains uneven across the Asia-Pacific. Many smaller economies and underserved networks have few monitoring points, leaving disruptions largely undocumented. Limited public reporting by governments and operators further reduces visibility into the status of critical infrastructure.

94 Leave a comment on paragraph 94 0 Against this backdrop, community-led monitoring can help address these gaps. In Taiwan, volunteers developed an open map of submarine cable faults, drawing public attention to infrastructure status. The government subsequently began publishing cable-status information in 2025. This demonstrated how community-generated evidence can contribute to greater public transparency.

95 Leave a comment on paragraph 95 0 In this context, status visibility requires expanding open measurement to underserved networks and making findings understandable to non-technical audiences. Governments should therefore publish timely information during disruptions, while regulators can use independently collected data to assess service quality. By achieving greater public understanding of infrastructure conditions support for accountability can be strengthened. 

96 Leave a comment on paragraph 96 0 2.2. Hidden dependencies

97 Leave a comment on paragraph 97 0 Internet resilience depends on more than the physical availability of cables and network connections. Essential services may rely on resources hosted overseas, including cloud infrastructure and external websites. Consequently, services can become inaccessible during international disruptions even when domestic networks remain operational.

98 Leave a comment on paragraph 98 0 Research in Taiwan highlighted the extent to which commonly used websites depend on overseas resources, raising questions about their reliability during international disconnection. These dependencies are particularly significant for economies reliant on a limited number of international connections.

99 Leave a comment on paragraph 99 0 Diversifying connectivity through cables, microwave and satellite links can reduce single points of failure. However, these alternatives also have limitations. Satellite connections may experience interference, while community mesh networks provide local communication but with restricted bandwidth and coverage.

100 Leave a comment on paragraph 100 0 Resilience planning must therefore examine the dependencies of essential digital services, not only the infrastructure carrying traffic. Establishing performance baselines for critical services and testing their behaviour during international disruptions can inform more effective contingency arrangements.

101 Leave a comment on paragraph 101 0 2.3 Sustainable Infrastructure Growth 

102 Leave a comment on paragraph 102 0 The expansion of digital infrastructure also raises environmental and governance challenges. AI-driven data centres are increasing demand for electricity and water, while some developments proceed without meaningful consultation with affected communities. At the same time, more efficient networks and digital services can reduce resource use in other sectors. Assessing sustainability therefore requires consideration of both the environmental costs and potential benefits of digital expansion.

103 Leave a comment on paragraph 103 0 The EcoInternet Index (EII), developed to compare Internet and climate-related indicators across Asia-Pacific economies, offers one approach to assessing these impacts. However, its use has largely centred on comparison rather than informing policy decisions. Efforts to make the data more frequent and actionable could strengthen its value for infrastructure planning.

104 Leave a comment on paragraph 104 0 Important measurement challenges remain, including how to account for AI-related demand and the environmental impact of hardware throughout its life cycle. Sustainability indicators should therefore inform investment decisions alongside technical performance and service accessibility.

105 Leave a comment on paragraph 105 0 Public participation is equally important. Communities affected by new infrastructure should be consulted, while civil society, researchers and youth can contribute to monitoring and evaluating its impacts. Open measurement tools and regional knowledge-sharing can strengthen local capacity to assess whether digital growth is environmentally sustainable and socially beneficial.

  1. 106 Leave a comment on paragraph 106 0
  2. Cyber Resilience, Trust & Cybersecurity governance and capacity

107 Leave a comment on paragraph 107 1 Cybersecurity is not only a technical or national security issue. Decisions about security also shape rights, distribute risk and determine who bears the cost of failure. Effective policy must protect systems without weakening fundamental rights. It must also assign responsibility to the actors best placed to reduce risk and equip institutions to enforce the rules. Across the Asia-Pacific, these challenges are intensified by ageing infrastructure, uneven regulatory capacity, workforce shortages and limited regional evidence.

108 Leave a comment on paragraph 108 0 A durable digital response combines proportional safeguards, lifecycle security, local expertise and regional cooperation. Smaller economies need particular support because they have fewer resources and less leverage over global firms.

109 Leave a comment on paragraph 109 0 3.1 Cyber Security, Cyber Rights and proportionality

110 Leave a comment on paragraph 110 0 Cybersecurity measures affect both system security and users’ rights. Their necessity, proportionality and wider effects should therefore be assessed before design and adoption. Encryption makes this tension clear. Governments may seek access through backdoors, client-side scanning, traceability rules or service shutdowns.

111 Leave a comment on paragraph 111 0 Reserved access cannot be made safe for legitimate authorities alone. Any weakness can be discovered or exploited by others. The WannaCry attack showed how state-developed tools can escape and be reused. Such proposals persist because of institutional incentives, public-safety framing and a commercial market for interception tools. Technical evidence alone rarely changes this dynamic.

112 Leave a comment on paragraph 112 0 Independent technical assessment and strong civil society oversight improves lawmaking. Legal safeguards should protect end-to-end encryption and these require transparent, evidence-based review.

113 Leave a comment on paragraph 113 0 Large platforms often defend privacy selectively. They resist state demands where they have leverage, but comply or withdraw where they do not. Users’ rights should not depend on a company’s market position. People in smaller economies are especially exposed because global firms may treat these markets as commercially marginal.Rules designed for large jurisdictions can also exceed the regulatory capacity of smaller economies. Regional models must account for these differences rather than export assumptions about resources and enforcement.

114 Leave a comment on paragraph 114 0 Cyber harms are unevenly distributed. Image-based abuse, doxing and gendered disinformation often target people because of their identity. An Inter-Parliamentary Union study found that 60% of surveyed women parliamentarians in the Asia-Pacific had faced such abuse. Women under 40 and women from minority groups were among those most affected. Other harms are indiscriminate but have unequal effects. Internet shutdowns hit people who rely on informal online income hardest, while health-data breaches can expose individuals to very different risks. Security measures can also cause harm. Laws against disinformation, for example, may be used to silence women journalists or other critical voices.

115 Leave a comment on paragraph 115 0 Cybersecurity measures should undergo independent review of their necessity, proportionality and differentiated impact. Remedies must be accessible to those most affected.  

116 Leave a comment on paragraph 116 0 3.2 Cyber Safety and Shared Responsibility

117 Leave a comment on paragraph 117 0 AI is shortening the time between the discovery of a vulnerability and its exploitation. As a result, deferred maintenance has become a strategic liability.  Actors use unpatched systems and end-of-life edge devices to reach telecommunications and critical infrastructure networks across Asia. An unpatched system can still be fixed, whereas an end-of-life system will never receive another update and needs to be replaced or isolated. Putting a new firewall in front of vulnerable equipment gives a false sense of security, and applying a patch does not prove that an intruder has left. Matching threat intelligence against an organisation’s own assets is recommended. AI can help with this, provided people check the results and make the appropriate response decisions.

118 Leave a comment on paragraph 118 0 Responsibility for equipment cyber vulnerability is fragmented. Manufacturers may have ended support, owners may not know their devices are vulnerable, and operators may see suspicious traffic with no access to the affected device. The issue is an IP address is where abuse becomes visible, not where it starts. The caution is against  treating registries and ISPs as the security police and arguments  are put forth for  moving responsibility upstream to manufacturers and supply chains. The EU Cyber Resilience Act, with its obligations on support periods, vulnerability management and visibility into software components is  a reference. However, any regional approach would need to reflect the Asia-Pacific’s diverse legal and institutional settings instead of assuming a single regulatory model. Public procurement is seen as a practical lever. Regulation in large markets is already having an effect in the region, with Taiwanese manufacturers starting to prepare for the Act as soon as its draft was published.

119 Leave a comment on paragraph 119 0 Market incentives alone are not  enough. Protections such as IPv6, RPKI and DNSSEC deliver value only when widely adopted, so early adopters carry costs without matching benefits. Taiwan’s IPv6 transition also showed that upgrading networks without lifecycle planning can add a further layer of technical debt. Rules for new products do not address the installed base of end-of-life devices, and several early-stage options are a matter of exploration. Vendors could coordinate extended support for widely used products that cannot be replaced immediately. Governments could attach security conditions to grants or direct their agencies to manage legacy technology. Companies could be required to report accumulated technical debt as part of their financial accountability.

120 Leave a comment on paragraph 120 0 Responsibility towards people is also uneven. National cybersecurity strategies tend to prioritise state systems, critical infrastructure and the financial sector. A preliminary analysis, covering 114 national strategies, found that only 34 refer to human rights and 14 to gender or women, mostly in relation to workforce participation. None set out a way to measure how threats or policies affect different groups. In practice, the burden of protection often falls on individuals, especially victims, more than on the institutions and platforms best placed to prevent harm.

121 Leave a comment on paragraph 121 0 3.3 Cyber Capacity and Cooperation 

122 Leave a comment on paragraph 122 0 Policies and powers fail without the capacity to carry them out. In Taiwan, police were given legal powers to use investigative technology without matching budgets for maintenance, access to vendors’ code and system logs, or adequate support for the officers operating the equipment. Forensic tools expired without updates and experienced staff left, weakening the reliability of digital evidence and, with it, due process. Securing and verifying the chain of custody was identified as the first priority where evidence systems may have been compromised. Budget constraints and rules on retiring public assets were also noted as barriers to replacing ageing equipment.

123 Leave a comment on paragraph 123 0 Incident response and reporting show similar gaps. Victims of technology-facilitated abuse often do not know how to preserve evidence, face stigma or fear being blamed, and find response mechanisms fragmented or missing. Access to remedies can also depend on income. Stakeholders identify capacity building for civil society, the technical community and policymakers as essential. ASEAN’s Women, Peace and Security agenda is a regional framework and offers one entry point. It is increasingly being linked to cybersecurity, and ASEAN guidance on national action plans identifying cybersecurity as an emerging challenge, although most plans have little about cyber issues.

124 Leave a comment on paragraph 124 0 In the Pacific, workforce constraints are acute. Populations are small, training is duplicated country by country, and certifications are not recognised across the region. Training and certification are paid for in foreign currency to offshore providers, tool licences are expensive, and labour mobility programmes draw skilled workers to other countries are critical to judge success by whether graduates can do the job, and to provide practical training, with relevant industry-aligned curricula and workplace experience. While cybersecurity skills are needed outside IT, including in human rights organisations, national computer emergency response teams are being established across the Pacific but need sustained support.  

125 Leave a comment on paragraph 125 0 Small economies cannot build every capability on their own. While every economy needs cybersecurity capabilities, not every economy needs every capability. Countries could specialise and share expertise, trainers and teaching materials adapted to local languages, and possibly computing resources.  

126 Leave a comment on paragraph 126 0 The region also lacks evidence about itself. South East Asia, the Pacific and smaller South Asian economies are largely missing from comparative research on encryption policy, data on how cyber threats and policies affect different communities is scarce, and Pacific research is often led from outside the region. An open regional dataset on encryption policy built with local researchers, and practical toolkits to help governments and civil society review national strategies is a strong proposition. What is encouraged is regional sharing of threat intelligence, end-of-life product information and procurement approaches, with clear roles, privacy safeguards and due process.  

  1. 127 Leave a comment on paragraph 127 0
  2. Governing Digital Public Infrastructure (DPI) and Personal Data

128 Leave a comment on paragraph 128 0 4.1 Choice and inclusion 

129 Leave a comment on paragraph 129 0 Digital public infrastructure (DPI), including digital identity, payment and data-sharing systems, increasingly determines how people access essential services and interact with the state. While these systems can improve efficiency and accessibility, their design may also create new forms of dependence and exclusion.

130 Leave a comment on paragraph 130 0 A central concern is whether digital identity systems preserve meaningful choice. In Malaysia, MyDigital ID is increasingly integrated into government services, while alternative means of access remain available. In Myanmar, biometric identification is compulsory for accessing certain documents and services, including passports and border passes. The comparison raises questions about when digital identification becomes effectively mandatory and whether consent remains meaningful when people lack practical alternatives.

131 Leave a comment on paragraph 131 0 Beyond the question of choice, the methods used to verify identity can themselves create barriers to accessing services through DPI. India’s Aadhaar system relies on fingerprint and iris recognition, which can create authentication difficulties for manual labourers, older people and persons with disabilities. Australia’s approach uses document verification and facial matching, avoiding some of these difficulties but creating barriers for individuals without smartphones or sufficient digital literacy. Neither approach eliminates exclusion entirely. Accessible, non-biometric or non-digital alternatives are therefore necessary to ensure that difficulties using a particular technology do not prevent people from accessing essential services.

132 Leave a comment on paragraph 132 0 4.2. Surveillance and rights

133 Leave a comment on paragraph 133 0 Digital identity and data infrastructures can facilitate state surveillance, when deployed without adequate legal safeguards and oversight. In Pakistan, proposals to restrict social media access for children under 16 raised concerns about age verification through government identification or facial recognition. Although child protection is a legitimate objective, linking identities to online accounts could increase the traceability of political expression and undermine anonymity, particularly in contexts where surveillance and cybercrime laws have been used against political dissent.

134 Leave a comment on paragraph 134 0 Experiences from Bangladesh and India reveal concerns about the relationship between surveillance powers and privacy protections. In Bangladesh, broad interception powers, inconsistent judicial outcomes and changes to oversight provisions across successive governments raises questions about the durability of privacy protections. In India, concerns include broad government exemptions under Section 17 of the country’s Digital Personal Data Protection (DPDP) Act, 2023, , executive oversight of interception orders and the reported use of surveillance technologies during protests.

135 Leave a comment on paragraph 135 0 These examples highlight the importance of examining how surveillance powers operate in practice. Legal authorisation alone is insufficient where technologies are deployed without transparent procurement, effective independent scrutiny or accessible avenues for challenging misuse. Oversight must address the surveillance systems themselves, including who controls them and how personal information may be used.

136 Leave a comment on paragraph 136 0 4.3. Safeguards and redress

137 Leave a comment on paragraph 137 0 The timing of legal safeguards is an important consideration in DPI governance. India and Australia illustrate contrasting approaches to the expansion of digital identity systems. India’s Aadhaar system was scaled before comprehensive legal protections were progressively established through legislation and judicial intervention. In contrast, Australia’s Digital ID Act 2024 introduced privacy safeguards, regulatory responsibilities and complaint mechanisms before broader expansion. However, Australia’s protections had yet to be tested at a scale comparable to Aadhaar. This example is meant to highlight that establishing safeguards before deployment can reduce the difficulty of correcting governance weaknesses after systems become widely used. Yet, it is acknowledged that effectiveness of these safeguards nevertheless depends on implementation and institutional capacity.

138 Leave a comment on paragraph 138 0 4.4. Public value and participation 

139 Leave a comment on paragraph 139 0 Transparency is essential to assessing whether DPI delivers its intended benefits. Often, public reporting on DPI uptake focuses on citizen registration figures without adequately examining service outcomes. High enrollment does not necessarily demonstrate improved access to services. DPI performance therefore should also be assessed through reliability, accessibility and the effectiveness of grievance mechanisms. Public reporting on these outcomes would enable citizens to evaluate whether digital infrastructure delivers its promised benefits.

140 Leave a comment on paragraph 140 0 The above task is compounded in the larger context of how meaningful participation should be a cornerstone of DPI governance. Youth, persons with disabilities and other affected communities should have opportunities to influence DPI design and governance, rather than being limited to consultation after decisions have been made. Public authorities and private technology providers must also be subject to appropriate accountability mechanisms.

141 Leave a comment on paragraph 141 0 These priorities become increasingly important as AI is integrated into public services. Weaknesses in identity and data governance may be amplified through AI-enabled profiling, exclusion and surveillance. Therefore, establishing privacy protections and independent oversight from the outset can help prevent these risks.

142 Leave a comment on paragraph 142 0 Trustworthy DPI governance ultimately depends on more than technological advancement. It requires systems that protect individual rights, provide effective remedies and deliver benefits that citizens can assess.

  1. 143 Leave a comment on paragraph 143 0
  2. Safety, rights, and resilience in contested digital environment

144 Leave a comment on paragraph 144 0 5.1 Digital rights for marginalized communities

145 Leave a comment on paragraph 145 0 Marginalized communities often face unique challenges to safe and meaningful digital participation. Women and queer people are two particular groups who experience increased soft barriers online, often relying on platforms and systems not designed with their safety in mind. Digital rights for these groups goes beyond pure access to the same systems as everyone else, and requires consideration of their actual experiences, including the risks they disproportionately face. 

146 Leave a comment on paragraph 146 0 Digital design choices must reflect the lived experience of marginalized groups, and how their experiences of harm differ. Some kinds of online risks which may seem gender-neutral, such as data breaches resulting in a leaked address or phone number, may have a greater impact on those already vulnerable to harassment, doxxing, or stalking. These kinds of online abuse, as well as others such as image-based sexual abuse (including deepfakes), already disproportionately affect these marginalized groups.

147 Leave a comment on paragraph 147 0 Members of vulnerable, stigmatised, and sometimes criminalised communities often rely on anonymity in order to participate fully online without fear of persecution or identity based digital harms. The inability to remain anonymous, control what information about them is shared, or otherwise customize the online experience to fit their needs, may therefore drive these communities away from meaningful digital participation and infringe upon their right to safety and inclusion.

148 Leave a comment on paragraph 148 0 These communities are highly vulnerable to online harms, but also rely on free expression and anonymity to participate fully, features that are often targeted with laws that purport to seek a reduction in digital risks. This tension creates a challenging environment for those most in need of digital rights protection online, and will require coordinated research, evaluation, and action to move towards a future where everyone can participate safely, freely, and without fear. 

149 Leave a comment on paragraph 149 0 5.2 Rights to Internet access and continuity

150 Leave a comment on paragraph 150 0 With growing reliance on the Internet for education, employment, essential services, community participation, and access to information, reliable Internet access is increasingly recognised as fundamental to the exercise of human rights. Meaningful access requires not only connectivity, but also continuity: people must be able to depend on an Internet that remains available, stable, and accessible when it is needed most.

151 Leave a comment on paragraph 151 0 Internet resilience is therefore not merely a technical requirement, but a human rights issue. Disruptions may result from infrastructure failures, natural disasters, or other technical challenges. However, Internet shutdowns are also increasingly imposed for political, security, and conflict-related purposes, restricting freedom of expression, access to information, peaceful assembly, and participation in public life.

152 Leave a comment on paragraph 152 0 The consequences of shutdowns are not experienced equally. Communities already facing social, economic, or political marginalisation often depend heavily on digital networks to access support, document abuses, maintain livelihoods, and communicate during crises. These communities may consequently experience the greatest harm when connectivity is restricted, particularly where alternative services and communications infrastructure are limited.

153 Leave a comment on paragraph 153 0 Protecting the right to Internet access and continuity requires coordinated and timely action. Governments, civil society, technical communities, legal experts, and regional networks must cooperate to strengthen resilient infrastructure and develop rapid-response mechanisms for documenting shutdowns, pursuing legal remedies, supporting affected communities, and coordinating advocacy across borders. Such measures can help ensure that Internet access remains reliable and is not weaponised to isolate communities or suppress fundamental rights.

154 Leave a comment on paragraph 154 0 Track 4: Universal and Meaningful Connectivity

155 Leave a comment on paragraph 155 0 In the Asia-Pacific region, the conversation around Universal and Meaningful Connectivity has definitively shifted from merely expanding baseline access to closely looking at the quality, resilience, safety, sustainability and ecological footprint of our digital infrastructure. No doubt, there exists comprehensive legal instruments and policy frameworks on paper and as best practices, the central bottleneck remains ground-level implementation, mostly in the last-mile. Providing internet access or hardware in rural, underserved, and crisis-affected areas without localized safety mechanisms and resilient physical infrastructure results in immediate community gains, but also introduces long-arching vulnerabilities rather than empowerment.

156 Leave a comment on paragraph 156 0

157 Leave a comment on paragraph 157 0 ​Discussions emphasized that meaningful connectivity must be rooted in sustainable, decentralized networks that can withstand severe disruptions—such as natural disasters, state-sponsored shutdowns, and geopolitical conflicts. Importantly, the narrative of digital expansion must urgently address the hidden environmental costs of routine technology use, particularly the massive resource consumption driven by emerging AI and large language models (LLMs), of late.

158 Leave a comment on paragraph 158 0 ​Effective internet governance requires a departure from rigid, top-down, adult-centric models. Children and youth—whose demographics form the majority of new internet users in the APAC region—must be elevated from passive beneficiaries to active co-designers of digital policy and co-architects of inclusive policy discussions. Achieving true meaningful connectivity requires updating regulatory frameworks that address emerging trends to support and facilitate community networks, enforcing Safety-by-Design principles on tech platforms, and deploying open, play-based capacity building that translates complex technical realities into actionable, local knowledge.

  1. 159 Leave a comment on paragraph 159 0
  2. Decentralized Resilience and Sustainable Digital Infrastructure

160 Leave a comment on paragraph 160 0 1.1 Demystifying Physical Infrastructure and Ecological Impact
Everyday digital activities—such as streaming video or executing AI queries—are often perceived as intangible, cloud-based processes. However, delivering data relies on a vast, physical global network infrastructure of routers, servers, transoceanic fiber-optic cables, and satellite links. There is the urgent need to demystify this physical reality for local communities. To engage communities effectively in policy discussions and digital governance and advocate for better governance and build local resilience, as a prerequisite, they need to understand how the internet physically works . ​Alongside this, there is a pressing need to address the environmental costs of digital technologies. Daily online habits and the growing use of AI tools require significant, measurable amounts of energy and water for cooling data centers, driving up carbon emissions. The environmental impact extends to device manufacturing, planned obsolescence, and raw mineral extraction. It is argued, to achieve environmentally sustainable meaningful connectivity that is sustainable, policies must support circular lifecycle models. This includes mandating renewable energy usage for infrastructure, promoting the repair and reuse of hardware to combat e-waste, and drawing on Indigenous knowledge systems to minimize the ecological footprint of digital expansion.

161 Leave a comment on paragraph 161 0 1.2 Building Resilient and Financially Self-Reliant Community Networks

162 Leave a comment on paragraph 162 0 Discussions veer around the fact that sustainable internet expansion requires hardware deployments capable of withstanding severe disruptions. Such frameworks must protect and facilitate decentralized, hybrid communication ecosystems—combining Low Earth Orbit (LEO) satellite tech, solar-powered access nodes, local Wi-Fi mesh networks, and offline distribution channels—to maintain essential lifelines during and after crises.

163 Leave a comment on paragraph 163 0
​However, relying solely on short-term philanthropy or energy-intensive infrastructure limits the long-term viability of these digital inclusion projects. To ensure survival of these solutions, for example, policies must enable community networks to adopt social-enterprise approaches; local networks can monetize excess bandwidth to fund public hubs like schools and health clinics, integrating financial self-reliance with community empowerment.

  1. 164 Leave a comment on paragraph 164 0
  2. Inclusive Governance and Grassroots Capacity Building

165 Leave a comment on paragraph 165 0 2.1 Modernizing Regulatory Frameworks for Inclusion
Traditional telecom licensing, rigid legislative structures, and top-down policy creation actively hinder digital expansion and fail to protect vulnerable demographics. Governance structures must evolve to keep pace with technological realities. Regulators are required and urged to streamline legacy licensing through “green lanes” (such as Open Access policies) and implement collaborative regulatory sandboxes. This allows communities to safely test local connectivity models before binding laws are drafted. Concurrently, policy frameworks must incorporate rights-based impact assessments—aligned with UN guidelines on children’s rights—and mandate the direct participation of youth and marginalized communities as key stakeholders in Internet governance.

166 Leave a comment on paragraph 166 0 2.2 Contextualized, Play-Based Learning and Youth Agency
It is widely observed and experienced that top-down, theoretical educational models fail to drive lasting engagement among young populations. Effective digital literacy programs must move beyond passive information delivery toward an action-oriented “Think, Reflect, Do” pedagogy. Emerging complex technical concepts—such as LLM tokenization, prediction algorithms, and data packet routing—can be effectively communicated through visual activities, hands-on simulations, and play-based tools. Leveraging successful models like career-connected learning cohorts and “Girls in Tech” programs can be of practical relevance. 

167 Leave a comment on paragraph 167 0 ​Meaningful inclusion calls for a multi-pronged approach. To this, governance resources and technical training must be open-access, peer-led, and translated into regional languages. By adapting these toolkits to distinct cultural and linguistic contexts, communities and youth can master technical concepts on their own terms, transforming from passive users into active operators and decision-makers in their local ecosystems.

  1. 168 Leave a comment on paragraph 168 0
  2. Moving Beyond Access: Safe Connectivity and Platform Accountability

169 Leave a comment on paragraph 169 0 3.1 Platform Design, Duty of Care, and Safety-by-Design
It is considered widely harmful to provide hardware and access without parallel safety frameworks. While digital literacy is frequently promoted as a solution, literacy alone cannot counteract platform features specifically designed to maximize screen time, capture attention, and drive engagement through addictive algorithms. Digital literacy must not be used by platforms or governments to offload the burden of safety onto children and parents. . Instead, Asia-Pacific internet governance policies must enforce strict Safety-by-Design principles, holding tech vendors, developers, and regulators legally accountable for the structural harms built into their digital tools. In alignment with proposed measures like the Asian Digital Safety Accord, platforms must be required to implement a “cognitive safe mode” (limiting rapid visual cuts and disabling engagement-driven recommendations by default for minors), prohibit targeted profiling of children, and undergo regular pacing audits.  

170 Leave a comment on paragraph 170 0 3.2 Protecting Communities in High-Vulnerability Contexts

171 Leave a comment on paragraph 171 0 Online safety frameworks must account for localized, crisis-affected environments. In extreme conditions—such as ethnic conflicts, digital displacement, or internet shutdowns—adolescents are disproportionately exposed to online violence, hate speech, and traumatic content without proper guidance. To protect these vulnerable groups, trusted local institutions must be recognized as core parts of the connectivity infrastructure including community libraries, ensuring youth retain access to safe, verified educational resources even during severe vulnerable contexts. To counter such vulnerabilities, grassroots initiatives like Community Access Networks (CAN) and Digital Peace Fellowships are vital. These mechanisms empower youth to engage in peer mentorship, mitigate viral hate, build digital confidence, and foster safe, constructive online engagement even during periods of severe social instability. 

  1. 172 Leave a comment on paragraph 172 0
  2. A Safer Internet for Children and the Role of Platform Design

173 Leave a comment on paragraph 173 0 4.1 Brevity bias and the pacing gap in child protection

174 Leave a comment on paragraph 174 0 The rapid growth of short-form video across the Asia-Pacific has moved the child online safety debate beyond what children see to how content is designed and delivered. Brevity bias, the tendency to prefer shorter and faster content over longer and more detailed material, is rewarded by the recommendation systems of short-video platforms, raising concern for children whose brains are still developing. Research project has analysed metadata from more than 180 short videos on children’s channels in Nepal, India, Bangladesh, Taiwan and Thailand. Average preferred content length was shortest in Nepal at 19.8 seconds, compared with 22 seconds in India, 25.4 seconds in Thailand and 28.1 seconds in parts of East Asia. Clips under 15 seconds drew about four times the engagement of videos longer than three minutes, and trending content tags overlapped by 92 percent across the countries studied. The research also projected a 42 percent decline in children’s attention spans by 2032.  

175 Leave a comment on paragraph 175 0 Discussions have identified a pacing gap in child protection law. Laws in the region, including those of Nepal and India, largely regulate what children can see, such as violent, obscene or abusive material, but not how content is delivered, including its length, speed and sequencing. Platform age limits are also easily bypassed when children give false information.  

176 Leave a comment on paragraph 176 0 4.2 A duty of care for platform design

177 Leave a comment on paragraph 177 0 Drawing on tort law, the area of law that covers civil wrongs and long-term damage, it is proposed a duty of care approach requiring platforms to take reasonable steps to prevent foreseeable harm, with harmful algorithmic pacing treated like a physical toxin. This would shift accountability from parents’ screen-time limits to platform design, cognitive sovereignty, a child’s right to neurological development free from manipulation, calls for  recognition in child welfare law. 

178 Leave a comment on paragraph 178 0 A proposed Asian Digital Safety Accord seeks to set out four measures, namely a cognitive safe mode limiting rapid visual cuts for users under 18; a boost for longer regional stories in recommendation feeds; regular pacing audits of platforms; and explanatory links on fast-moving protest clips. The cognitive safe mode is seen as the most achievable today. Australia’s emerging duty of care legislation, which follows its ban on social media for children under 16, is cited as a related development, alongside the view that raising baseline safety may protect children more than service bans.

179 Leave a comment on paragraph 179 0 Accountability is  shared among platforms, governments, researchers, civil society, educators, parents and young people. With users now creating short-form content themselves, the question is  whether responsibility lies with creators or with the platforms that amplify their content. Parents need support and a place in policy discussions, since educating children alone leaves them unprotected. Further suggestions include switching off engagement-driven recommendations by default for minors and prohibiting targeted profiling of children, while ensuring that protective measures respect children’s privacy, freedom of expression and access to information.

Page 22

Source: https://comment.rigf.asia/?replytopara=152