WS #144 Bridging the Digital Divide Language Inclusion As a Pillar

Summary

This discussion focused on bridging the digital divide through language inclusion as a fundamental pillar of digital rights and internet governance. The panel, moderated by Ram Mohan and featuring experts from organizations like ICANN, Unicode Consortium, and various domain registries, explored how linguistic diversity remains an overlooked barrier to digital inclusion despite global efforts to expand internet access. With over 7,000 languages spoken worldwide, the dominance of a few major languages online prevents billions of users from fully participating in digital spaces.

Panelists highlighted numerous challenges, including limited multilingual content availability, poor machine translations, cultural irrelevance in literal translations, and bias in AI systems toward dominant languages. They emphasized that these barriers create a vicious cycle where younger generations lose fluency in ancestral languages, further weakening the user base for language inclusivity efforts. The discussion revealed that while technical solutions like Unicode character encoding and internationalized domain names exist, the primary obstacles are awareness, implementation, and adoption rather than technological limitations.

A central theme emerged around the need for a fundamental paradigm shift from “English first” to “multilingual by design” approaches. Speakers stressed that language accessibility should be treated as a core component of digital infrastructure rather than an afterthought or localization add-on. They advocated for universal acceptance by design, where technology serves languages rather than forcing languages to adapt to technology.

The panel discussed various practical solutions, including government incentives, educational initiatives, procurement requirements that prioritize multilingual support, and public-private partnerships. They highlighted successful examples like hackathons teaching students to build bilingual websites and community grants supporting indigenous language preservation. The discussion concluded with calls for stronger multi-stakeholder collaboration through initiatives like the Coalition on Digital Impact (CODI) to coordinate efforts and accelerate progress toward a truly multilingual internet that ensures meaningful connectivity for all users.

Keypoints

## Major Discussion Points:

– **Language barriers as a fundamental obstacle to digital inclusion**: Despite global efforts to expand internet access, linguistic diversity remains an overlooked barrier, with over 7,000 languages worldwide but only a few dominant languages represented online, preventing billions from fully participating in digital spaces.

– **The need for a paradigm shift from “English-first” to “multilingual-by-design”**: Panelists emphasized moving away from treating multilingual support as an afterthought or localization add-on, instead making it a foundational element of digital infrastructure from the beginning of development processes.

– **Technical solutions and standards already exist but lack widespread implementation**: The underlying technology for multilingual internet experiences (Unicode, internationalized domain names, universal acceptance) has been largely developed, but the challenge lies in awareness, adoption, and overcoming inertia in existing systems.

– **Multi-stakeholder collaboration is essential for progress**: Success requires coordinated efforts across governments (through policy and incentives), private sector (through inclusive design practices), civil society organizations, educational institutions, and language communities themselves.

– **Cultural preservation and economic opportunities**: Multilingual internet access is crucial not only for preserving endangered languages and cultural heritage but also for enabling meaningful economic participation and safer online experiences for non-English speaking populations.

## Overall Purpose:

The discussion aimed to explore how language accessibility can be integrated into internet governance and digital rights frameworks, examining both the challenges preventing linguistic inclusion online and practical solutions for creating universal access and acceptance across all languages.

## Overall Tone:

The discussion maintained a consistently collaborative and solution-oriented tone throughout. While panelists acknowledged significant challenges and barriers, the conversation remained optimistic and forward-looking, with participants sharing concrete examples of successful initiatives and practical next steps. The tone was professional yet passionate, reflecting the speakers’ deep commitment to digital inclusion. The discussion concluded on an encouraging note with calls to action and invitations for broader participation in addressing these challenges collectively.

Speakers

**Speakers from the provided list:**

– **Ram Mohan** – Chief Strategy Officer of Identity Digital (domain name registry company), Chair of the Coalition on Digital Impact (CODI)

– **Christian Daswon** – Executive Director of the Internet Infrastructure Coalition, Co-founder and co-facilitator of CODI, Online moderator for the session

– **Theresa Swinehart** – ICANN representative, oversees policy implementation, DNS abuse programs, Universal Acceptance and internationalized domain names work

– **Toral Cowieson** – CEO of the Unicode Consortium (nonprofit, open source, open standards body focused on ensuring communication in all languages on all devices)

– **Jennifer Chung** – Vice President of Policy for DotAsia Organization (DotAsia top-level domain registry operator)

– **Sophie Mitchell** – Chief Communications Officer at DOTAU (Registry Operator for Australia)

– **Manal Ismail** – National Telecom and Regulatory Authority of Egypt, European Sector member working on internationalized domain names and universal acceptance

– **Audience** – Multiple audience members who asked questions during the session

**Additional speakers:**

– **Mohammed Abdul Haq Onu** – Secretary General of Bangladesh Internet Governance Forum

– **Gabriel** – Audience member who asked about website implementation challenges

– **Elisabeth Carrera** – Wikimedia Norway representative

– **Roberto Gaetano** – Individual user representative who commented on multilingualism improvements

# Comprehensive Discussion Report: Bridging the Digital Divide Through Language Inclusion

## Executive Summary

This comprehensive discussion examined language inclusion as a fundamental pillar of digital rights and internet governance, bringing together experts from organizations including ICANN, Unicode Consortium, and various domain registries. The panel, moderated by Ram Mohan (Chief Strategy Officer of Identity Digital and Chair of the Coalition on Digital Impact) and Christian Daswon (Executive Director of the Internet Infrastructure Coalition), explored how linguistic diversity remains a critical yet overlooked barrier to meaningful digital participation despite global efforts to expand internet access.

Ram Mohan opened by highlighting that while over 7,000 languages are spoken worldwide, only a handful dominate online spaces, preventing billions of users from fully participating in digital environments. Panelists emphasized that this creates not merely inconvenience but genuine safety risks, economic exclusion, and cultural erosion. The conversation maintained a consistently collaborative and solution-oriented tone throughout, with participants sharing concrete examples of successful initiatives while acknowledging significant systemic challenges that require coordinated multi-stakeholder responses.

## Key Participants and Their Perspectives

The discussion featured diverse expertise across technical standards, policy implementation, and regional perspectives. **Theresa Swinehart** from ICANN provided insights into policy implementation and DNS abuse programs, emphasizing that no single organization can solve multilingual internet challenges alone. **Toral Cowieson**, CEO of the Unicode Consortium, offered technical perspectives on character encoding while acknowledging limitations in supporting oral-only languages. **Jennifer Chung** from DotAsia Organisation consistently advocated for paradigm shifts from “English first” to “multilingual first” thinking, while **Sophie Mitchell** from DOTAU shared Australia’s experiences with Indigenous language preservation. **Manal Ismail** from Egypt’s National Telecom Regulatory Authority contributed policy and regulatory perspectives, particularly around government incentives and procurement approaches.

Audience participation enriched the discussion significantly, with **Mohammed Abdul Haq Onu** from Bangladesh highlighting Universal Acceptance challenges despite 250 million Bangla speakers and three years of Universal Acceptance Day events at the Bangladesh Internet Governance Forum focusing on Bangla language. **Elisabeth Carrera** from Wikimedia Norway shared optimistic news about AI systems learning minority languages, noting that Northern Sami Wikipedia sees “88% bot/AI traffic vs 12% human traffic, with AI desperately trying to learn the language.” **Roberto Gaetano** provided a particularly thought-provoking critique of the community’s own practices, using an Animal Farm metaphor to highlight how even multilingual advocates often default to English-dominant behaviors.

## Fundamental Barriers to Digital Linguistic Inclusion

### The Scale of Exclusion

The discussion established that linguistic barriers represent a massive yet underappreciated form of digital exclusion. This exclusion extends beyond basic access to what Sophie Mitchell termed “meaningful connectivity” – the ability to use the internet for essential services like healthcare, education, and government services rather than merely basic functions like social media.

Jennifer Chung introduced a critical safety dimension through a personal example about her elderly father and healthcare appointments, explaining that language barriers create cybersecurity vulnerabilities as people cannot identify suspicious content or phishing attempts in unfamiliar languages. This reframing elevated the discussion from convenience to safety, demonstrating how linguistic exclusion creates tangible harm.

### Technical Infrastructure Challenges

Toral Cowieson provided crucial insights into the invisible infrastructure of digital privilege, explaining that “for those of us in majority languages there are many things that we can take for granted. So for example, can I exchange currency, date and time format, time zone adjustments, usage of numerical symbols or decimals and commas separators used appropriately.” This comment grounded the discussion in concrete, granular examples of digital exclusion that most people never consider.

The technical challenges extend beyond basic character encoding. Theresa Swinehart highlighted Universal Acceptance issues that prevent proper functionality of internationalized domain names across applications and systems. She provided a simplified definition: “the resolvability of one’s experience with one’s address or with one’s website or with one’s domain name and for that to interface through the different applications.”

An audience member (Gabriel) raised practical implementation barriers, noting that font rendering and screen reader accessibility remain obstacles for website creators wanting to implement their native languages, even when they have the motivation to do so. Toral mentioned the issue of “Tofus” – little squares that appear when characters don’t render properly – illustrating ongoing technical challenges.

### Cultural and Generational Impact

The discussion revealed how digital linguistic exclusion creates vicious cycles of cultural erosion. Jennifer Chung presented a stark statistic: “90% of the world’s languages might become extinct at the end of this century. That’s like, that is a big loss, not in terms of how we communicate, but really, it’s human culture.” This reframed the entire discussion from a technical accessibility issue to a cultural preservation crisis.

Sophie Mitchell provided specific context from Australia, where the population includes 8.5 million people born overseas out of 28 million total, and “about half of its 250 Indigenous language groups” have been lost due to colonization, with many remaining languages endangered and originally oral rather than written. Manal Ismail explained how limited multilingual content availability and poor machine translations create cycles where younger generations lose fluency in ancestral languages, further weakening the user base for language inclusivity efforts.

## The Paradigm Shift: From English-First to Multilingual-by-Design

Jennifer Chung introduced what became the central theme of the discussion: “The one thing that will accelerate it the most is actually a paradigm shift to thinking about looking at it as multilingual first as opposed to English first. To look at it by, you know, multilingual by design, universal acceptance, UA by design.” This concept was immediately adopted and referenced by multiple other panelists throughout the remainder of the discussion.

Manal Ismail provided an equally powerful articulation of the required change: “So instead of making languages change technology, we need to start making technology serve languages.” This phrase captured the power dynamic at the heart of the digital divide, showing how current systems force linguistic communities to adapt to technology rather than technology adapting to serve diverse communities.

The speakers demonstrated strong consensus that language inclusion must be treated as foundational digital infrastructure rather than an afterthought. This represents a fundamental shift from treating multilingual support as a localization add-on to making it a core design principle from the beginning.

## Technical Solutions and Implementation Challenges

### Existing Infrastructure and Capabilities

The discussion revealed that many technical foundations for multilingual internet experiences already exist but suffer from implementation and adoption challenges. Toral Cowieson explained that Unicode encoding has addressed core technical limitations, but challenges remain in engaging language communities and ensuring full user experience. He provided a specific example of how the spaceship emoji renders differently in various tools, noting that Canva works correctly because “Unicode was in its technical stack.”

Theresa Swinehart outlined ICANN’s support for internationalized domain names and Universal Acceptance through technical standards, education programs, and partnerships with universities. She highlighted a successful hackathon in Bahrain with 60 students creating bilingual Arabic-English websites. However, she emphasized that the primary obstacles are awareness, implementation, and adoption rather than technological limitations.

### Gaps in Current Approaches

Despite technical progress, significant gaps remain in supporting the full spectrum of linguistic diversity. Toral Cowieson acknowledged that the Unicode Consortium focuses on written languages, stating: “that is not something that the Unicode Consortium is addressing at this point because of the work still to be done with the written scripts.” This creates a fundamental limitation for many Indigenous and minority languages that were traditionally oral.

Elisabeth Carrera from Wikimedia Norway highlighted the need for quality text and data to be available under open licenses, noting that AI systems need access to quality multilingual data, but much existing content in minority languages isn’t available under open licenses for training purposes. Christian Daswon referenced the Creative Commons CC Signals program as addressing some LLM scraping concerns.

## Policy and Governance Approaches

### Government Incentives and Regulatory Frameworks

The speakers demonstrated strong agreement on the potential for government policy to drive multilingual adoption, though they emphasized different mechanisms. Manal Ismail advocated for comprehensive regulatory approaches, suggesting that governments should integrate multilingual support into digital accessibility laws, procurement requirements, and national digital transformation strategies. She proposed both incentive-based approaches like tax breaks and grants, and regulatory mandates making language support prerequisites for government contracts.

Jennifer Chung focused on de-incentivizing non-compliance through procurement processes and requiring Universal Acceptance readiness reporting in tenders. This approach uses market mechanisms rather than direct regulation to encourage adoption. Theresa Swinehart preferred soft enforcement mechanisms including education system requirements, government agency leadership by example, and procurement policy preferences.

### Leading by Example

A recurring theme was the importance of government agencies and educational institutions demonstrating multilingual technologies in their own operations. Theresa Swinehart emphasized that leading by example through government agencies using multilingual technologies creates both awareness and market demand. This approach avoids regulatory mandates while creating practical demonstrations of feasibility and benefits.

The discussion also highlighted the role of educational integration through university modules, hackathons, and curriculum development to create awareness among the next generation. This educational approach addresses both technical capacity building and cultural change needed to support multilingual internet development.

## Multi-Stakeholder Collaboration and Coordination

### The Essential Role of Partnership

There was unanimous agreement among speakers that the complexity of creating a multilingual internet requires coordinated efforts across multiple stakeholder groups. Theresa Swinehart stated definitively that “no single organization can solve multilingual internet challenges alone; success requires coordinated multi-stakeholder approach.” Toral Cowieson reinforced this by noting “we can only be successful together.”

Christian Daswon introduced the Coalition on Digital Impact (CODI) as a coordinating mechanism, explaining that it “aims to bring together diverse stakeholders and coordinate existing efforts rather than duplicate work.” This approach recognizes that numerous organizations are already working on various aspects of multilingual internet development, but coordination and communication need improvement.

### Clarifying Roles and Responsibilities

Toral Cowieson highlighted that strengthening existing cross-stakeholder work requires clarity on roles and responsibilities, better communication, and accountability measures. The discussion revealed that while many organizations have relevant expertise and programs, language communities often lack clear pathways to engage and contribute to solutions.

The speakers agreed on working simultaneously on supply and demand sides – creating user awareness while building technical capabilities. This dual approach recognizes that lack of demand reduces supply incentives and vice versa, requiring coordinated efforts to break the cycle.

## Economic and Social Benefits

### Business Case for Multilingual Internet

The discussion revealed strong consensus on the business case for multilingual internet access. Theresa Swinehart explained that multilingual internet access creates economic opportunities by connecting businesses with previously unreachable consumer populations. Sophie Mitchell noted that capacity building and digital skills training in native languages increases economic participation and civic engagement.

However, Christian Daswon identified a critical gap in available data, asking “How can we gather data that shows if people build tools for language inclusion at the last mile, recipients will use the Internet more?” He explained that while the benefits seem obvious, hard data is needed to convince businesses to commit time and resources to language inclusion tools.

### Social and Cultural Value

Beyond economic benefits, speakers emphasized the social value of multilingual internet access. Sophie Mitchell highlighted how digital inclusion enables access to essential services like healthcare, education, and government services in users’ native languages. Jennifer Chung positioned language preservation as maintaining cultural identity and human heritage while enabling meaningful digital participation.

The discussion also revealed how multilingual internet access contributes to civic engagement and democratic participation. When people can access government services, educational resources, and civic information in their native languages, they become more engaged citizens and participants in democratic processes.

## Artificial Intelligence: Opportunity and Opportunity

### AI’s Positive Potential

Elisabeth Carrera from Wikimedia Norway provided optimistic news about AI’s role in minority language development, saying “I just thought I’d add a little bit of optimism” before sharing that Northern Sami Wikipedia sees significant AI traffic with systems “desperately trying to learn Northern Sami.” She framed this as hopeful: “So I think there’s hope.”

Sophie Mitchell explored the potential of AI translation technology, asking thoughtful questions about “whether technology will help solve it” while acknowledging both opportunities and challenges in AI-assisted multilingual internet development.

### Data Inequality Concerns

Christian Daswon framed AI challenges in terms of data inequality, noting that “in the era of advancing AI, when people are using the internet directly with LLMs. There are these data-rich languages and many, many, many data-poor languages.” This creates new forms of digital divide where AI systems amplify advantages for languages with abundant training data while marginalizing those without.

The discussion highlighted the need for quality text and data to be available under open licenses for AI training, but also recognized language communities’ concerns about giving up ownership rights. This tension between open data needs and community control represents an ongoing challenge requiring careful balance.

## Practical Implementation Strategies

### Individual and Organizational Actions

The discussion concluded with concrete action items for participants. Theresa Swinehart suggested that individuals can contribute through personal networks and professional connections to raise awareness and create opportunities. Each participant was encouraged to contact two people in their networks to raise awareness about multilingual internet opportunities.

Organizations were encouraged to integrate language awareness into existing educational curricula and professional programs rather than creating entirely new initiatives. This approach builds on existing infrastructure while expanding scope to include linguistic inclusion considerations.

### Systemic Changes Required

Beyond individual actions, speakers identified systemic changes needed across multiple sectors. Government agencies and organizations were urged to lead by example by implementing multilingual technologies in their own operations. Stakeholders were encouraged to treat language inclusion as a requirement in procurement processes and tender evaluations.

The discussion emphasized working with existing successful initiatives rather than starting new competing programs. Toral Cowieson noted the importance of “connecting existing successful initiatives rather than starting new programs, and making it easier for language communities to participate.”

## Unresolved Challenges and Future Directions

### Technical and Implementation Gaps

Several significant challenges remain unresolved. The question of how to effectively support oral-only languages that lack written representation in digital systems represents a fundamental gap that current technical approaches cannot address. Font rendering and accessibility tool limitations continue to prevent website creators from implementing native languages, even when they have the motivation to do so.

### Data and Evidence Needs

Christian Daswon identified critical data gaps that prevent effective business cases for multilingual internet investment. He emphasized the need to gather evidence showing that language inclusion tools increase internet usage. Without this evidence, convincing businesses to invest in language inclusion tools remains challenging.

The discussion also revealed the need for better data collection from last-mile connectivity providers to understand what tools and support they need for language inclusion. This direct feedback from practitioners could inform more effective solution development.

### Cultural and Behavioral Change

Roberto Gaetano provided perhaps the most challenging critique, using an Animal Farm metaphor to highlight how even multilingual advocates often default to English-dominant behaviors: “We behave in a way that I would call the syndrome of animal farm, in the sense that we all say, okay, all languages are equal, but then in practice we behave as if we have one language that is more equal than the others.” He noted this in the context of interpretation services and panel behavior.

This comment challenged the audience to examine their own practices, not just advocate for policy changes, highlighting the need for cultural change within the very communities promoting linguistic inclusion.

## Conclusions and Path Forward

The discussion demonstrated remarkable consensus among diverse stakeholders on both the urgency of addressing linguistic barriers to digital inclusion and the fundamental approaches needed. Jennifer Chung concluded by emphasizing: “It’s absolutely a language justice movement. And again, repeating multilingual first and UA by design.”

The Coalition on Digital Impact (CODI) emerged as a concrete mechanism for ongoing coordination, with Christian Daswon inviting continued engagement through CODI postcards and the call to action. The commitment to survey last-mile connectivity providers and gather evidence of language inclusion benefits represents practical next steps toward building stronger business cases for investment.

However, Ram Mohan raised a closing concern about AI potentially leading to “one language to rule them all,” highlighting ongoing challenges in the AI era. The discussion revealed the depth of systemic change required, moving from “English first” to “multilingual by design” thinking requires not just technical solutions but fundamental shifts in how digital systems are conceived, developed, and implemented.

The path forward requires sustained multi-stakeholder collaboration, with clear roles and responsibilities, adequate funding for both technical development and community engagement, and ongoing commitment to treating language inclusion as foundational digital infrastructure rather than optional enhancement. The discussion provided both the conceptual framework and practical starting points for this essential work, while acknowledging the significant challenges that remain in creating a truly multilingual internet that serves all of humanity’s linguistic diversity. The reference to UNESCO’s International Decade of Indigenous Languages provides additional context for the urgency of this work.

Ram Mohan: Good morning and good day. My name is Ram Mohan and welcome to the session Bridging the Digital Divide, Language Inclusion as a Pillar. Despite global efforts to expand Internet access, linguistic diversity remains an overlooked barrier to digital inclusion. With over 7,000 languages spoken worldwide, the dominance of a few major languages that are online prevents billions of users from fully participating in digital spaces. Today’s workshop will explore how language accessibility can be integrated into Internet governance as well as digital rights frameworks and what connectivity strategies can ensure universal access and universal acceptance. What we have here is a panel of experts and leaders who’ve been working in the multilingual domain name and multilingual space and we will have both a bit of a policy discussion as well as a conversation about actual use cases that demonstrate both the need as well as the ability to solve. for the actual problems of digital inclusion. So let me begin first with some introductions. Let me start with myself. My name is Ram Mohan. I am the Chief Strategy Officer of Identity Digital, but which is a domain name registry company. But I’m here in addition to that as the chair of an exciting and newly formed group, the Coalition on Digital Impact or CODI in short. And we’re really pleased to be here and we invite you to become a part of CODI by going to codi.global. Let me turn to my online moderator, Christian, ask him to introduce himself and then carry on from there.

Christian Daswon: Happy to do so. I am the online moderator for today. My name is Christian Dawson. I’m Executive Director of the Internet Infrastructure Coalition. We’re a tech trade association. We have been working for over a decade on issues involving universal acceptance. And I’m proud to be a co-founder and co-facilitator of the CODI effort that we’ll talk a little bit about today.

Ram Mohan: Theresa.

Theresa Swinehart: Yes. Ah, ah. There we go. This is a press the button one. Okay, fantastic. Let me try that again. Theresa Swinehart, I’m with ICANN. I oversee our work with regards to policy implementation and various other aspects on DNS abuse and some of the programs that we have that are across collaboration with the community and the board as well. In one area in particular here is our work on not only the next round of top level domains which will afford opportunities. but also Universal Acceptance and our work around internationalized domain names. So very excited to be here and to look forward to the discussion.

Ram Mohan: Thank you. Thank you so much. Let me turn to you, Toral.

Toral Cowieson: Great. Good morning. Good afternoon. Good evening, everyone. Thank you, Ram and Cody, for this invitation. I’m Toral Cowieson. I’m the CEO of the Unicode Consortium. Unicode is a nonprofit, open source, open standards body whose mission is to ensure that everyone can communicate on all devices in their own languages. Digital inclusion starts with Unicode, starts with the work of Unicode, which includes character encoding.

Jennifer Chung: Thank you, Ram and Cody, for the invitation. My name is Jennifer Chung. I’m Vice President of Policy for DotAsia Organization. We are the DotAsia top-level domain registry operator. In terms of internationalized domain names, we have done quite a lot of work and, of course, we support Universal Acceptance and looking forward to the discussion. Back to you, Ram.

Ram Mohan: Thank you so much. Sophie, let me come to you.

Sophie Mitchell: Thanks very much, Ram. Hi, everyone. I’m really pleased to join you from here in Australia. I’m Sophie Mitchell. I’m the Chief Communications Officer at the DOTAU, so the Registry Operator for Australia. And yeah, OUTER is really committed to enhancing digital inclusion and improving the utility of the internet or the DOTAU for the benefit of all Australians. We do quite a bit of work in this space through our public benefit and grants program, which we may touch on in the course of this evening. So, thanks, Ram. Back to you.

Ram Mohan: Thank you so much. And Manal.

Manal Ismail: Thank you, Ram. Hello, everyone. My name is Manal Ismail. I’m with the National Telecom and Regulatory Authority of Egypt and I’ve been a member of the European Sector for some 5 years. working on introduction of internationalized domain names and universal acceptance efforts and now progressing multilingual internet experiences for quite a number of years now and I look forward to our discussion today. Thank you Ram, back to you.

Ram Mohan: Thank you Manal, so we are looking to have a discussion here in a couple of parts but the first part I thought that might be useful to illustrate the nature of the issues that are facing us and the kinds of not just challenges but also some potential solutions that might be there. We thought it might be instructive to ask folks who are actually doing the work of inclusion and to have them share what they are finding from the ground. So Toral, shall I turn to you and ask you for what you’re seeing in your role and in your organization’s role in this area.

Toral Cowieson: Certainly, thank you so much. So when we’re looking at digital inclusion, for those of us in majority languages there are many things that we can take for granted. So for example, can I exchange currency, date and time format, time zone adjustments, usage of numerical symbols or decimals and commas separators used appropriately. If I am using an application, does the country name appear appropriately. And so those are things that we take for granted but that’s not necessarily the case for those in digitally disadvantaged language communities. And I think that’s where the bulk of the work has to be done, Ram, when you mentioned the 7,000 languages. The majority of languages, the characters of those languages are encoded, but it’s all of those other things that we take for granted. Can I search on my emoji in the language that I use? So there’s still a lot to do, and I think even with IDNs, we know that from a content perspective, that full representation and that full user interface is not there.

Ram Mohan: Thank you, Toral. Jen, what do you see?

Jennifer Chung: I guess in terms of multilingual internet, I think that is really the foreground and the background and all through the entire thread of what we’re looking at. I’m going to take a half step back, because when you look online and you’re searching online, if your language is not English, is not your native language, you’re not really searching in that language. I will take out my phone, I’ll be searching in Chinese, I’ll be speaking in Cantonese to my phone as well. So that’s the first thing. And of course, the content also still, right, online, almost 50% of that content is still in English. And of course, even though English is most widely spoken language, in terms of population, in terms of breadth, it is not the native language or first language or even second or third languages for a lot of people who use the internet and those who, of course, are yet to come online. I think it’s really important to remember the other part of the need to actually think about why we need this multilingual internet is if we don’t have this, exactly what you said, all these languages will go away. I think there was a statistic somewhere that said, you know, 90% of the world’s languages might become extinct at the end of this century. That’s like, that is a big loss, not in terms of how we communicate, but really, it’s human culture. So just kind of just putting this out there as a first kind of demonstrating the need and something that we really do need to start looking at and doing something about.

Ram Mohan: Thank you, Jen. Sophie, what has your experience been in the regions that you and your organization have been involved in?

Sophie Mitchell: Thanks, Ram. So, yeah, just taking it back to here in Australia, I think Australia is one of the greatest countries in the world. I, of course, would say that. But we have about 30% of our population are actually born overseas. So they don’t just come from parents who were born overseas, they are themselves born overseas. That’s in a population of 28 million, it’s about 8.5 million people or thereabouts that are born overseas. Some of those come with English language skills, but many, of course, do not. So that’s our migrant population. But, of course, Australia was a colonised country. So first Australians who’ve been living here for many thousands of years actually have 250 language groups, or indeed they did when the British arrived and settled Australia. Unfortunately, through the process of colonisation over the last 200 plus years, that we’ve lost about half of those Indigenous language groups and many of those are endangered today. So certainly echo what Jen was saying, and that is an issue here in Australia is capturing those endangered languages. Because, of course, the other problem we have in capturing them is that they weren’t written languages, they were oral languages and told through a culture of storytelling. And, of course, that culture was impacted adversely by colonisation because Indigenous families were, children were separated from their families and moved on to reservations, as has occurred in many colonial countries. And so a lot of that knowledge was lost. So really, we’re starting from a step back in trying to find those languages, preserve those languages. And then, of course, then we can address some of the digital inequities that we see with Indigenous communities. And I’m sure we’ll go on to talk about those this evening. Thanks, Ram. .

Ram Mohan: Thank you, Sophie. So, as you heard from, you know, Jen, what you’re saying, as well as what you just shared with us, Sophie, the challenges are not just in representing languages, but in some cases, representing languages that have no written representation. Toral, how does that work, you know, from the perspective of the Unicode Consortium?

Toral Cowieson: That’s certainly a very real issue, and in fact, Unicode’s work is really around written languages, so that is not something that the Unicode Consortium is addressing at this point because of the work still to be done with the written scripts. So, I’m curious to hear how others are addressing that, but that’s just the reality of Unicode’s mission.

Ram Mohan: Manal, what is your perspective on other challenges that may exist when it comes to, you know, getting the internet to be more multilingual?

Manal Ismail: Thank you, Ram, and from a government perspective, of course, truly multilingual internet is crucial for digital inclusion. It’s essential for the continued growing of online population and necessary to have the next billion users connect meaningfully to the internet. As already mentioned, we have already one-third of the world’s population still offline, and it was quite alarming to read in the Fact of the Future that progress on most of the SDGs is either moving too slowly or has regressed below the 2015 baseline. So, I think we need to think about the reasons and how to make sure technology is really serving humanity in that respect. And while digital inclusion really promises global information access, language remains a fundamental axis of inequality and a barrier that’s commonly overlooked. It hinders equitable, meaningful digital participation. It restricts individuals’ access to critical services, limits economic opportunities, and as already Jen mentioned, it erodes cultural identity. And we have a long list of challenges here. The limited availability of online multilingual content, poor quality or absence of machine translations, cultural irrelevance resulting from literal translations without really cultural localization. As you already mentioned, some platforms and softwares prioritize a small subset of global dominant languages. And this also touches on the bias in AI and natural language processing. High cost of building robust digital tools, making it economically unfeasible. Sometimes really deprioritizing language accessibility in favor of faster and broader market scalability. And even when it comes to trainings, they are often delivered in dominant languages resulting in digital literacy gaps. And all this unfortunately results in a vicious cycle with younger generations not being fluent in ancestral languages, thus weakening even more the user base for potential language inclusivity. I’ll leave it at this. Back to you, Ram. Thank you.

Ram Mohan: Manal, that is powerful what you just said. And you listed not just the some of the challenges, but also some of the consequences of not addressing these challenges. Theresa, what are your points of view on the policy gaps, as well as some of the technical challenges that we’re seeing? And what is your organization doing to help bridge those? And I’d also be interested to hear your views on what you think are the remaining inequities even after the work that is being done.

Theresa Swinehart: Those are a lot of questions. We have a couple hours, right? So there’s a couple factors, and I think what I’m really quite struck by in these conversations is the internet itself. And within our context of what ICANN does, the addressing space is an incredibly powerful medium. And its goal is to be unfragmented and enable more people to come online. And yet at the same time, we are seeing a saddening erosion of use of languages around that. And so from that, I think we need to cross this barrier of, yes, the internet was originally designed in the context of the use of Roman character sets, and that’s just a historical factor. Had it developed in a different part of the world, it would have been different. But how do we now create the right awareness that we can bridge into these areas? And so from ICANN’s perspective, we’re engaged in quite a few different things. One is the work around internationalized domain names. And so from a technical standpoint, enabling the use with language tables and various others. on the technical interface from that standpoint. And there’s a lot of material around that and colleagues working on that. And with that, you know, how does one not only have one’s email or anything of that sort, but also to the right or to the left of the dot, if I can put it that way, in the context of Arabic scripts. So, working with many communities there. And then partnering with many other organizations with regards to awareness of universal acceptance. Now, the term universal acceptance may sound a little bit odd, but what it actually means is the resolvability of one’s experience with one’s address or with one’s website or with one’s domain name and for that to interface through the different applications and enable a communication. Now, that’s a very simplified description of it. And I think the engineers would probably just pull me aside and say that’s, you know, there’s a lot more to this. But the point is that we have a demand for languages. You know, UNESCO is running a decade of indigenous languages and this is quite important for preserving languages, whether they’re written or oral. But we now need to do an awareness that there is a demand and the supply needs to meet the demand. And that disconnect is one that’s fundamentally important. It is feasible to have one’s entire experience online as one would expect offline. But we need to now take that next step to make that feasible. I’ll just do one more thing and then I’ll hand it over. So we have been engaged with many others. We have a partnership, MOU, with UNESCO, with many other organizations and working with them for awareness, but also have a universal acceptance day that is really looking at creating awareness. at local levels about the opportunities that can occur if one works towards this. So we have a lot of work to do, but I’m hopeful. And for oral languages, that is the content that can then be online, and it’s fundamentally important around that. So a few things here and there, and I’m happy to add on later.

Ram Mohan: Thank you. Thank you, Theresa. So, Sophie, I’m just going to come to you and ask, what kinds of incentive-based approaches have you and your organisation thought of as effective ways for the adoption of digital content and services in local languages?

Sophie Mitchell: Thanks, Ram. I think one of the things we’re focused on here, and I think Theresa summed it up really nicely then when she said we need supply to meet demand. And so part of the problem we see here is that Indigenous languages aren’t widely spoken, and we also have a really quite a big skills gap in Australia in our Indigenous population. So that has been mapped. The government’s been doing some work since the early 2020s on the Indigenous digital inclusion, and we do have in Australia an Indigenous digital inclusion index. So this gap is actually mapped, and we’ve just had funding to continue that index over the next three years. So there’s an access gap, so Indigenous access, there’s an ability gap, so that’s that skills gap, and also an affordability gap, because we do see in Australia that our Indigenous population are more likely to live in remote areas and therefore less likely to have access to high-paying jobs in those areas. And so ARDA, as I mentioned earlier, does through its public benefit fund have a community grants program through which it funds a number of initiatives. And the initiatives we’ve really tried to fund over recent years have been about supporting both recording Indigenous languages so that they can be used, and also increasing the skills of Indigenous communities, so their digital skills, so that they can participate in the digital economy, and upskill so that they, you know, may get jobs and I guess also can participate. in civic society and access government services as well, because that’s another area we see gaps. People who lack those digital skills tend to use the internet in a more simplistic way. So often for social connection, so using social media or for telephone calls, but they don’t extend that use for other purposes. So it can be quite exclusionary and they’re things we’re looking to try and improve so we can increase that economic participation and grow Australia’s digital economy. Thanks, Ram.

Ram Mohan: Fantastic. So Jen, Sophie just brought in this concept of meaningful connectivity, right? It’s not enough to simply be connected and stream and use for social media, but make that connectivity more meaningful. What do you see in the Asia Pacific region in terms of meaningful connectivity? Many governments and other folks are spending lots of money connecting populations, but what challenges exist to make connectivity meaningful to get populations to actually achieve agency online?

Jennifer Chung: Thanks, Ram. I’m gonna pick up on the last word. You mentioned agency is absolutely the most important part because after you connect the populations who are previously unconnected or underserved, they might not know how to use technology. They might not know what kind of ways they can search. For example, like I said, they wouldn’t know that they could search on with their native language if all the content they see right now is in a different language that they don’t understand. Capacity building is definitely the key thing here, but then targeted to actually bringing them online to understand what they can do. There could be e-services, there could be education. and the different parts of capacity building must really direct to the populations that are most vulnerable for them to understand. One thing that I really want to point out is multilingual internet is not only about preserving language. In fact, having it in the thinking that multilingual internet actually makes it safer. And you know, for example, I like to take my father, who’s a bit elderly now, if I tell him that he needs to do certain things on this website to make an appointment for his health checkup, he would most likely be like, I won’t want to do this if I can’t read this. I don’t know. I don’t understand. I have to find someone else to explain to me all these different procedures. But if everything is able to be searched, navigated content wise, also in his native language, he will feel a lot better. Another thing I want to point out is, you know, we talk all about, you know, different bad actors online try to spoof these things, but actually understanding the language makes it much easier for perhaps populations who don’t have English as a first language to first understand, oh, this actually looks a little bit strange to me and understand that, you know, I’m fortunate to be able to understand that when, you know, I see some spoof email or something like that, I won’t click on it because that looks strange to me and I can read the English. But for someone who absolutely don’t understand these characters at all, that could be a huge mystery. And I think understanding that having this is actually makes for a safer Internet is something that we need to think about.

Ram Mohan: Thanks Jen. Christian, you’ve been helping work on this area for quite a long while. And you know, recently you helped co-found Kodi. What is happening there to help address some of these challenges?

Christian Daswon: Thanks Ram. I’m really glad that Jen brought up cyber security. I think that’s a very important topic. There are a number of different reasons why focusing on digital inclusion is so important right now, and that is certainly one of them. There are, surfacing all of the reasons why it is important to act on this at the moment, it’s one of the goals of CODI as an initiative. The initiative is focused on attempting to bring together all the different parts of the community that are going to need to collaborate together in order to make a difference, to move this forward. One of the fundamental things that we sort of have to realize is that the underlying technology in many ways has been complete and has been complete for some time. What we’re dealing with is a situation where the awareness is required because we are fighting inertia. And there’s a timeliness now in the era of advancing AI, when people are using the internet directly with LLMs. There are these data-rich languages and many, many, many data-poor languages. And so you get into these situations where we need to advance all of the arguments for people to put the effort in to go ahead and take the technologies that already exist and advance them towards multilingualism. And so we need to bring the arguments about cybersecurity, about the importance of inclusion so that they can be part of the digital economy and grow the economy. But also, as we were saying, because we can’t lose the stories and the histories that exist from these cultures. There are many reasons why we learn history in order to make sure that we don’t repeat it, but also so that we can live harmoniously in societies and not advance towards war and disharmony. and all of these different arguments are the things that we need to bring to bear. So we’ve got a number of initiatives that we’re pulling together with CODI. Some are simple. One of them that we’re advancing at the moment is we’re starting to try to figure out how to gather data that will show that if people build tools for language inclusion at the last mile, it will be more likely that the individuals who are recipients of those tools will use the Internet more. It’s quite obvious that that would be the case, but in order for businesses to go ahead and commit time and resources, quite obvious isn’t good enough. They actually need hard data, and there aren’t really a lot of studies that show these sorts of things. So that’s just one example of some of the many projects that we’re going to be putting together and advancing with CODI, and we’re hoping that you join us.

Ram Mohan: Thanks so much. Manal, in terms of government initiatives to support language accessibility in digital spaces, what has been the experience with your government?

Manal Ismail: Thank you. Thank you very much. When the Internet fails to speak your language, it’s not just inconvenient, it becomes a barrier to health, education, and livelihood. As Sophie hinted, governments have an important and powerful role in encouraging the development and adoption of multilingual digital content in order to foster digital inclusion. This can be a mix of both regulatory approaches but also incentive-based approaches in order to balance enforcement. with support of frameworks and capacity building. Regulatory approaches may include mandating multilingual access for key digital platforms and services, especially in public sectors, expanding digital accessibility laws and regulations to explicitly include language accessibility alongside disabilities, and making multilingual support a prerequisite for public procurement contracts or national projects. As for the incentives, these may also include offering targeted grants and funding for projects that develop digital tools, content or platforms in underrepresented languages, providing tax breaks or reductions or subsidies even to companies that integrate inclusive language practices, and forming public-private partnerships also with technology firms and or universities to build and maintain inclusive digital ecosystems. It would be also good to establish a recognition or some sort of certification program for digital inclusion and digital language inclusion in order to recognize platforms that meet government requirements in that respect. Finally, also investing in capacity building and training programs for developers and translators as needed. So, in a nutshell, in order to ensure that inclusive language support is not an afterthought but a core pillar of national digital transformation strategies, Effective government strategies should integrate regulatory mandates to ensure compliance, incentive schemes to ensure investment and innovation, and collaborative frameworks to ensure sustainability and community participation. Thank you Ram, back to you.

Ram Mohan: Thank you Manal. Toral, what are the fundamental technical limitations or design paradigms that have historically contributed to the Internet’s bias linguistically towards just a few dominant languages?

Toral Cowieson: So I think, you know, the technical limitations have been, at least from a Unicode perspective, have been addressed on our side, but where the challenge is, is really engaging the language communities themselves for the remaining work to be done. And that’s where, you know, when I talked about kind of that full immersed experience, that’s where the limitation is. And, you know, part of it is on the Unicode side is people don’t know that there is work to be done, and how do we effectively equip and empower those communities to be part of the solution to ensure that solution is in place. So that’s really, that’s really the issue. And, you know, our member organizations have tended to be the big tech companies, and originally the priorities were the majority languages. But now they are also realizing that it doesn’t, it serves their better interests as well to make sure these other communities are online. And so we do see organizations like, you know, certainly Google, Meta, Airbnb, where they have access to native speakers. and they’re bringing those individuals into that work for that full experience. So, you know, from a kind of a core technology perspective, it’s just getting language communities.

Ram Mohan: Thank you, Toral. I’m going to turn to now, you know, we’ve spent some time looking at the challenges, identifying some of the barriers that exist. Let’s turn our attention to practical things that we can actually do. So, I’m actually going to turn to each of the folks here on the panel and I’m going to ask you this question. Looking at the landscape of existing efforts, what is one area where you see significant potential for accelerating progress towards a multilingual internet? So, let me ask that and turn to you, Jen, first.

Jennifer Chung: The one thing that will accelerate it the most, I think the one thing that will accelerate it the most is actually a paradigm shift to thinking about looking at it as multilingual first as opposed to English first. To look at it by, you know, multilingual by design, universal acceptance, UA by design. That actually sets us up for success because when we have that mentality, all of the stakeholder groups would be, the governments will understand exactly what Manal says. You know, this is part of national strategy. It’s not an afterthought. For those who are coding, they will know this is what it is that we learn. This is how we do it instead of doing it the previous way. In terms of, you know, private sector too, that’s also the incentive. So, that is really the mind shift that’s the one thing that will accelerate it.

Theresa Swinehart: I think it’s spot on to have a paradigm shift, that one should have an expectation that one’s experience online is the same as offline. I think there’s a couple areas where one can look at that. One is around education and educating and creating modules in coursework, whether it’s around policy or around technical areas. We run a program where we are working with universities on specific modules to have awareness and education around these areas. We just recently ran a hackathon in Bahrain, which was incredible to watch 60 students have a task of coming up with a website and how to have that be able to work in both Arabic and in English. So I think there’s some very pragmatic, very specific things to do. I think there’s also some other areas around procurement and that there is just an expectation that one is going to in the future need to supply things in this kind of way. And to add on to that, I think that there’s also opportunities as we’ve talked about. We open the next round next year with applicant support for specific categories to create the opportunity to have what one wants to the right or the left of the dot in one’s language. But it’s fundamentally that there is a paradigm shift, as has been said, about the expectation is, as a user, have the experience in the online digital world as one would in the offline world. And the benefit is not only societal. It’s not only the opportunity to engage with a hospital or a university or education or provide that information and retain cultural norms and cultural values. One’s home, you know, that is anchored in who we are as individuals and how we communicate with each other. There’s also economic opportunities from a business perspective. You know, there’s entire populations that have no consumer interest potentially in whatever somebody is providing because they don’t have a connection because of the language barrier. So, it’s not about economics only, it’s about society and affording that ability for everybody to have that experience in the Internet. And so, I think we have many, many, many things to do. From ICANN’s perspective, we’re doing some of them very specific and partnering with others because nobody can do this alone.

Ram Mohan: Thanks, Theresa. Sophie, one area where you see significant potential for accelerating progress towards a multilingual Internet?

Theresa Swinehart: Thanks, Ram. I find it really difficult because I think both Jen and Theresa there have hit on two really key elements. I think, yeah, people don’t participate because they’re, in many cases, because they’re excluded so they can’t, and that mindset shift isn’t there. I’ve spent a lot of time thinking, since I was invited to speak on this panel, about whether technology will help solve it. So, will AI translation actually accelerate it or be part of the problem? Because do we worry about a multilingual Internet if you don’t have to, or if you’ve just got those translation skills there? So, I think that that’s not really a mindset shift, but I do think it will be a very big shift and I hope accelerate the availability of the Internet for those who are currently excluded. But I do still think it’s that upskilling. Even if you have the language skills, certainly I know we find in migrants in Australia, many arrive without English skills, but they actually, many of them don’t have great skills in their own native languages either. So, if you’re illiterate, that’s probably a barrier to using the Internet in any language. So, we spend a lot of time focusing on multilingual abilities, but I think it’s also, sorry, technology, but I think it’s also upskilling people with digital skills in whatever language they have as well.

Ram Mohan: Christian.

Christian Daswon: For as long as I’ve been coming to IGF, we’ve been talking about connecting the next billion. We are, we are focusing our attention on how we can connect the unconnected. And there are lots of stakeholders who are doing that, putting efforts into last mile connectivity. We’ve spent time here talking about meaningful connectivity. And earlier today, you asked Jen what she thought it would take to add language along with access devices and training into the definition of meaningful connectivity everywhere. What I would really like to do is to go out individually to all those stakeholders to have a survey, to actually go to the people that are connecting the unconnected with last mile and ask them en masse, try to gather up that information and then disseminate it. Because all of us then need to take what they are saying they are missing, what they are telling us they are missing, and we need to figure out how we can collectively build those tools. So what I want most is I want direct information for the people that aren’t making it happen now and for them to tell us how we can make it happen.

Ram Mohan: Toral?

Toral Cowieson: Yeah, thank you. Those are hard answers to follow. I don’t know what to add. So I actually want to build off of something that Theresa said, that no one can do this alone. And I think, you know, the one key thing is strengthening the work that’s already happening across stakeholders. You know, when we think about all the stakeholders who are required, NGOs, governments, tech companies, standards organizations, language communities, and really taking a step back and organizing ourselves, kind of clarity around the roles and responsibilities that each of us have. in this multilingual internet, and then communicating that effectively, because there’s so much work. We talked about UNESCO’s International Decade of Indigenous Languages, for example, and all the other initiatives, ICANN’s initiatives, for example, and how do we bring those together in a way that we can make it clear for how language communities can be engaged? Because I think we make it so hard, and it’s hard to know, all of us have so many things coming at us. How do we make it easier? And then how do we engage and empower those communities to be, again, part of that solution? So, you know, my thoughts are really around our work across stakeholders and the responsibility that we have to one another, to the language communities, and how do we hold ourselves accountable to success? You know, like Christian said, we’ve been talking about that next billion for a long time, and how are we actually demonstrating that we’re bringing that to closure in some way? So, thanks, great question.

Ram Mohan: Thank you. Manal?

Manal Ismail: Thank you, Ram, and I truly believe that the most significant mindset shift needed from all stakeholders, as we mentioned, across the internet ecosystem, is to treat language equity not as a localization add-on that comes as an afterthought, but as a core driver of global meaningful participation and a foundational layer of digital infrastructure. I mean, same as connectivity, as cyber security. And accordingly, rather than thinking of linguistic plurality as a challenge that we need to fix, and sometimes by adding translations later. or by trying to find workarounds, we all know how IDNs were introduced and keep it if there is demand. I think we need to think of multilingual by design, as Jen also mentioned, as an integral to inclusive, scalable and ethical digital infrastructure innovation. We also need conscious and deliberate, maybe national initiatives that involve all stakeholders, like has been mentioned also the importance of collaboration in that respect. Such initiatives could be government-supported, publicly-funded, open-source-based, something that could provide high-quality language resources, tools or APIs for underrepresented languages. So instead of making languages change technology, we need to start making technology serve languages. Thank you, Ram. Back to you.

Ram Mohan: Thank you, Manal. That’s a really powerful way to encapsulate what we’re trying to do. Technology should be serving languages rather than the other way around. So let me ask a question. You know, we have here in our panel, we have diverse perspectives, technical perspectives, civil society, industry, government. So, Christian, let me start with you. How can these diverse stakeholders collaborate and coordinate so that we can amplify their impact in promoting linguistic inclusion in digital policies?

Christian Daswon: I think we have a lot of people with a lot of good intentions and a lot of us don’t know exactly what it is we need to do to put those good intentions to active good. There are lots of programs that do active good in here. We’ve heard about some of the active good programs that ICANN has been doing for a long time in this area. One of the things that we’re doing with COTI is we’re trying to find all the people that are doing active good. We’re trying to bring them into a sort of single coordinating stream where people can talk to one another about things that are going on and raise programs that need to get done. I think there just needs to be a lot more talking and a lot more organization around all of the efforts that are going on so that we can centralize them and surface other ways to help that will get people to be able to raise their hands and say, well, that’s the kind of work that I can do. I’ve been looking for a way to contribute to these efforts. I’m ready.

Ram Mohan: Theresa, what do you think? How can we better coordinate and collaborate among the various stakeholders? If the goal is to have language as a pillar of digital inclusion strategies worldwide, what can we do better?

Theresa Swinehart: I think not be shy to pursue ideas. By that I mean different organizations may be engaged, I’ll just give some examples, with educational modules around internet-related issues or digital inclusion-related aspects or how certain things function. Are there opportunities there to add on additional modules around language awareness, around internationalized domain names or universal acceptance or how to build that in? I think sometimes looking at things in isolation and our own work in isolation, we lose sight of the opportunity that for somebody who’s involved in a, let’s say, policy track of education for international relations, let’s say, by example. It might not make sense to create awareness that there is actually the opportunity to have one’s own language online and what that would entail. But a chapter or two in a book that is part of that sets off that awareness. In a business school, awareness that one actually, from a business standpoint, has a population that could very well be interested in using their own languages for educational institutions or for medical professions, that one serves a population. Many of us, when we go to our medical facility, it may say, if you need somebody to help interpret this, please contact this number and we can provide these languages. Where is that online? Where is that source of the click down? So I think that there’s, we tend to look at either our areas of work in isolation or our sectors in isolation, and we need to do what we can to help connect the dots. And in those dots, what are the opportunities, what are the tools in the toolbox for different sectors, different organizations to fully serve a population that is not only speaking with Roman character sets, by example, from a technical standpoint, but is speaking so many different languages and some of them are written in oral. And how does one then go to serve them? So it’s a big task, but I would say if each of us walked away today and said, here are the two things I’m going to follow up on and the two people I know in wherever it might be, I’m going to contact and say, hey, did you know? it could go a long way.

Ram Mohan: Jen, what mechanisms are missing or need strengthening when we look at collaboration and coordination?

Jennifer Chung: I think both Christian and Theresa have actually hit on a lot of these points, and Toral as well, and when she mentioned, you know, you can’t go at it alone. Missing things still, I mean, just filling in the gaps, I think what it is. Theresa’s last call for action really is personal for all of us here at the table and those in the room as well. Making that connection, connecting the two dots, understanding that each of the organizations, DotAsia, of course, we have a lot of work that we do in IDNs, and we do a lot of work in capacity building, but in the act of doing all these things, we come across different organizations also in the region that can help us with these efforts, especially when we’re looking at CODI, you know, connecting good things that people are doing and allowing them to do them together. So again, I mean, the answer is the same thing, filling in the gaps in the way that we know how. I think that is what is missing and that really takes a collective effort.

Ram Mohan: Thank you so much. So let me turn to you, Manal, with a different question. So we’ve been talking here about, you know, the principles of linguistic diversity and multilingual by design. So how can these principles become more explicitly embedded into digital rights frameworks and more embedded in governance discussions?

Manal Ismail: Thank you, Ram. Can you hear me? Yeah, so, as I mentioned earlier, This could be extended in laws and regulations that already support disabilities and this could be also part of such laws and regulations. And have this expanding the digital accessibility laws and regulations to explicitly include language accessibility alongside disabilities. And also through the regulatory and approaches we’ve mentioned and the incentives that could be done as well. And I believe also as everyone mentioned it’s a multi-stakeholder thing and it’s also as a special characteristic that we need to work on the supply and demand side at the same time. As some have already mentioned, even the demand side they don’t know they can have this service. So they see everything in Latin, in English or whatever dominant languages and they don’t know that this luxury is even possible. And on the other side, the supply side, they don’t see it visible from a business point of view. So we need to work on wearing the demand side and triggering the supply side at the same time. And maybe governments can help with national initiatives, help with government support or public funds. And lastly, also connecting the dots and making sure that we’re all pushing in the same direction as everyone else mentioned. I think one good thing about the Universal Acceptance Global Day that we can learn from, it started globally and now it’s connecting with on-the-ground national activities, which I think is a good thing that we can also follow when we talk about a really multi-lingual and digital future.

Ram Mohan: Thanks Manal. Let’s go to getting questions and really an interaction with the audience. Let’s start online, Christian.

Christian Daswon: Sure, we do have one question in the chat. Feel free if you’re in the chat to add others and we’ll also welcome people to come to the mics as well. You’ll go into the queue next. Fantastic. A question from Nicholas Fumarelli. How can countries accelerate IDN compliance and readiness in the light that these measures are not advancing from year to year at the same pace of other adoptions, such as RPKI validation enabled networks or DNSSEC deployment? What can be alternatives other than law enforcement on this topic? I’ll read the rest. We’re actually having a fragmentation of languages due to this. Some efforts from UASG to reach popular software owners, messaging apps, among others. But how can, for example, local chapters help? Do you have any examples of soft enforcement for UA readiness? To note, UA is also part of UNESCO Rome X framework. Would anybody on our panel like to take this?

Ram Mohan: I think Theresa is probably closest to the response to that question. Christian, I think there were several questions in there. So, let’s get to just one thing that perhaps…

Christian Daswon: So, let’s see. Let’s focus in on the idea soft enforcement for UA readiness. How can on the idea soft enforcement for UA readiness? How can countries accelerate IDN compliance and readiness with soft enforcement mechanisms?

Theresa Swinehart: So, I think that there’s many different ways, and we saw efforts around with IPv6 at different points. I think that there’s soft ways to do it are through the education system and the next generation through the schools and the opportunities to both create awareness but maybe require that some of the digital engagement is done through the use of internationalized domain names or universal acceptance tools and work with the technical community in order to make sure that the schools have that capacity. That creates awareness for the next generation, which is essential if we want to actually preserve languages moving forward. Maybe it’s also soft ways through requirements of medical facilities or government social facilities that are providing to their citizens that the government agencies themselves look to have those mechanisms put into place and work with providing that for their citizens in their languages. Many countries have multiple languages, and then they have indigenous languages as well that they could serve. So, I think leading by example is essential that if one wants to actually have this resonate, one has to lead by example around that. I think that there’s other ways within the context of the private sector. Maybe if one’s procurement policy may suggest that one would encourage the use of that, and if not, to be willing to help to figure out how different agencies could do that or different bidders could do that. So, I think that there’s many, many different ways to build it into one’s own effort, but part of it is walking the talk and leading by example. I think that’s an important fundamental step around that.

Ram Mohan: Thank you. Jen?

Jennifer Chung: I don’t know if I like the phrase soft enforcement, but really I think we’re looking at incentivizing, especially when you’re looking at industry. I think a lot of, you know, registries, registrars, we are incentivized to be able to do this, you know, in our own internal systems as well. And then I want to pick up on something that Theresa mentioned. Not so much, again, not so much self-enforcement, but de-incentivizing. I think Manal also touched on it when she mentioned, you know, tax incentives, tax breaks, putting it, requiring it in the procurement process in kind of stages where you first, you know, prioritize those contractors that do have this. And then finally, you know, in the end, you know, this would become a requirement. And then including the requirement of reporting of UA readiness in tenders, I think that, you know, combined would have the effect that I think the answer was looking for. Yeah.

Ram Mohan: Thank you, Jen. Let’s go to the gentleman here.

Audience: Thank you. This is Mohammed Abdul Haq Onu, or everybody call Onu. I’m a Secretary General of Bangladesh Internet Governance Forum. Myself with Bangladesh Internet Governance Forum actively organizing Universal Acceptance Day event for the last three years with a special focus on Bangla language. I truly welcome this discussion on language inclusion. Using the digital device is not only about infrastructure, it is about ensuring every community can access the internet in their own language. In Bangladesh, our efforts to promote UA and local script adoption have shown that digital inclusion become meaningful when people can use technology in their native language. Despite the challenges spoken by over 200 Bangla languages, 250 million people still face challenges in full UA readiness. I urge global stakeholders to treat language inclusion as a core digital right to invest in technical and policy framework. that support meaningful internet access in multilingual access also. Also challenges meaningful internet connectivity, especially rural areas. Thank you.

Ram Mohan: Thank you so much for your thoughtful intervention, appreciate it. Let’s take another question from the audience there.

Audience: Hello, my name is Gabriel and I have a question regarding implementation. So for someone wanting to implement a website in their own language, it’s often that you would get into trouble with, for example, font rendering not displaying it properly or the font not having support. And when you get into other accessibility things like screen readers, if even those are mandated, they might not have support for that language. So my question is then how do you ensure that the underlying tools that people are using to build websites are accessible to these languages so that other people who are educators or bloggers can make websites in their own languages without having to have deep knowledge about how all these things about font rendering or technology works. Thank you.

Ram Mohan: Thank you so much. I think, Toral, that is adjacent to some of the core things that you’re doing, so perhaps you want to take that.

Toral Cowieson: Yeah, it’s a great question. I’ll give you an example. I was recently working with a designer on some stickers or something, and we were using, let’s say, the spaceship emoji, and what are the different languages in which you could search for that. And we put it into one tool. I’m not going to say which organization. And we were getting a lot of Tofus, right, little squares with just nothing that was useful. And then we put it, I said, let me try it in Canva. And I put that same block of information into Canva and everything rendered correctly. And what that told me is that Canva had built Unicode, had Unicode in its technical stack. Technology stack. And what was different about how Canva was approaching it is their products are built to be world ready. So Canva is not a Unicode member, so I’m just, let me just put that out there. But it really depends on what tools you’re using. And if Unicode is in the tech stack, that will make it easier for your website development to happen. So it’s really finding the right tool and not being frustrated that something isn’t working the way that you were hoping that it would. But there are solutions that are out there. And if Unicode is in the tech stack, that can accelerate the work.

Ram Mohan: And Christian, I wonder whether that idea that Toral was just speaking about and what the gentleman was identifying as a real need may be one of the kinds of initiatives that might be useful on a global scale.

Christian Daswon: Well, I think that’s exactly why the organization that we’re trying to build is focused on listening. Because identifying where there are gaps is an important thing that we need to be doing. But also where there is effective work that needs to be connected with the people who need to know that it’s being done is also a vital tool that needs to be done as well.

Ram Mohan: Thank you so much. I’ll go to the lady there on my right.

Audience: Hi, my name is Elisabeth Carrera, and I’m with Wikimedia Norway. And one of the many things we do to do with language diversity is we support the community that creates and curates Northern Sami Wikipedia. So we’re sort of the movement behind Wikipedia. And I just thought I’d add a little bit of optimism when it comes to tech and language. diversity. Well, what we’re seeing now, you mentioned we should make technology serve languages. I think technology is quite keen to go bilingual, full force. With our encyclopedia in Northern Sami, currently, human traffic is about 12%. And Northern Sami is a tiny language. The remaining traffic is bots, spiders, large language models, desperately trying to learn Northern Sami. So I think there’s hope. One of the obstacles we’re facing with small languages is just the same that we faced when we first started gaining like public access, general access to the internet here in Norway in the 1990s. There isn’t enough text, quality text and quality data available. So frustratingly, in Northern Sami, in this case, text is available. It’s just not available under open licenses. So it’s not there to be scraped. So if you want tech to learn multilingual approaches quickly, we need to make available what we have. So that’s the first step. And it’s a very elementary step that we can easily make. And even public entities don’t realize how much of a difference they can make by making their data sets available under open licenses. So just a small piece of optimism, we can easily do this.

Ram Mohan: Thank you so much, Christian.

Christian Daswon: I’ve got a couple of thoughts. First of all, one of the things that we have been doing as we’ve been trying to organize CODI is a movement that we are trying to start in this area is talking to a number of groups who are doing important work in this space. And I want to say that the work that Wikimedia Foundation and Wikipedia are doing is phenomenal. They’re doing incredible things when it comes to language and in the areas where we are talking about the challenges that are coming because of AI. They’re doing it not just in languages, but in cultural contexts, right? And the things that they’re doing there are incredibly impressive. I think that you’re right, that open needs to be an important component of what it is we’re talking about. Of course, open has its own challenges in the era of LLMs because large data scraping is a challenge for a lot of people who are designed to have open environments. Yesterday, I joined the kickoff call for a Creative Commons program called CC Signals that’s trying to address some of the concerns that people have when it comes to scraping in open environments for LLMs. And that’s an area that we’re going to try to connect CODI to. I encourage everybody to take a look at the efforts that are happening there because I agree with you, it’s very important and hopeful.

Ram Mohan: Thank you. Toral?

Toral Cowieson: Great, thank you for that positive news there. I just wanted to comment on this notion of open. And this is where we’ve hit some concerns from language communities that by granting rights that they are giving up something. And that is an ongoing challenge. One of the distinctions that we’ve been working on as we’ve been talking with language communities is this open license is the only way for full participation and access for other users. And being really clear that by licensing that you are granting licensing but you’re not giving up. You still retain ownership. And I think that distinction is clear as we can make it. And I’ve had to work with our general counsel on this a bit because it can be very easy to get deep into the legalese and I think that frightens and scares off language communities, rightfully so. and just raises some huge concerns, but that open license piece is a huge step towards access.

Ram Mohan: Thank you. We have another question from the audience on my left.

Audience: Yes, Roberto Gaetano, Uralo Individual Users. This is not really a question, it’s rather a comment or a contribution on the question, what can we do to improve multilingualism? I think that I agree completely with what has been said, that we need a paradigm shift. We have really a cultural problem with this, and this is something that is now, that we are subject to this particular mentality. We behave in a way that I would call the syndrome of animal farm, in the sense that we all say, okay, all languages are equal, but then in practice we behave as if we have one language that is more equal than the others. And let me make an example. In a lot of meetings we have interpretation, also here in the main room at ICANN meetings we have several situations in which we have interpretation. And the situation is more often than not that the panel consistently speaks in English. And the interpretation is used by people who speak English. in the audience to understand, to listen in their own language. And just rarely there’s a question from the audience in a different language and then maybe somebody from the panel as a courtesy replies in that language. But why can’t we establish the rule, the behavior that when we have interpretation and your language is one of the interpreted languages, why don’t you speak in your own language or in the language that you are more comfortable with than in English? I think that this cultural change is done like every cultural change by little steps. We don’t need huge changes, but just make the people think that there are more languages. This will not achieve the 7,000 languages having same status over the internet, but at least we can start moving slowly in that direction.

Ram Mohan: Thank you so much. In many ways, I think, especially with the advent of large language models and artificial intelligence engines, we’re quickly moving to a path where we may end up having one language to rule them all, rather than the diversity of languages that really represents the world around us. So, thank you so much. I’m going to ask each of the panelists one sentence that sums up what you want to provide to our audience as a takeaway, but only one sentence. Let me start with you, Sophie.

Sophie Mitchell: Oh, Ram, that’s unfair. I’m still ruminating on the question from the last gentleman about animal farm. So, I almost want my one overarching thought to be, let’s not be animal farm. and let’s ensure that all languages are equal. But I think, am I allowed one more? Sorry Ram, I’m already breaking your rule.

Ram Mohan: Please go ahead.

Sophie Mitchell: I was just gonna say it’s such a multifaceted issue that I really think it does require a multi-stakeholder response. And I think if we start quantifying the benefits for people, then we might be able to see better unification and unified action. And I’d love to see that.

Ram Mohan: Thank you. Toral, one sentence.

Toral Cowieson: We can only be successful together.

Ram Mohan: Thank you. Manal.

Manal Ismail: We need to all continue working on the supply and demand side together. Multilingual by design, not an afterthought. Thank you.

Ram Mohan: Thank you. Jen.

Jennifer Chung: It’s absolutely a language justice movement. And again, repeating multilingual first and UA by design.

Ram Mohan: Thank you, Theresa.

Theresa Swinehart: Lead by example and connect everybody to change this.

Ram Mohan: And last word, Christian.

Christian Daswon: My sentence is, there are Cody postcards on the corner table over there. You should grab one and let’s do more together than we can apart.

Ram Mohan: Thank you so much. That concludes this session. Thank you.

Agreement points

Paradigm shift from English-first to multilingual-by-design approach

Speakers

– Jennifer Chung
– Manal Ismail
– Theresa Swinehart

Arguments

Need fundamental shift from “English first” to “multilingual first” thinking, with Universal Acceptance designed in from the beginning rather than added later

Technology should serve languages rather than forcing languages to adapt to technology limitations

Language inclusion should be treated as foundational digital infrastructure like connectivity and cybersecurity, not as localization add-on

Users should expect same online experience as offline experience, with language accessibility as core expectation rather than luxury

Summary

All speakers agree that the fundamental approach must shift from treating multilingual support as an afterthought to making it a core design principle from the beginning, treating language accessibility as foundational infrastructure rather than optional add-on

Topics

Digital standards | Multilingualism | Human rights principles

Multi-stakeholder collaboration is essential for success

Speakers

– Toral Cowieson
– Theresa Swinehart
– Christian Daswon
– Sophie Mitchell

Arguments

We can only be successful together

No single organization can solve multilingual internet challenges alone; success requires coordinated multi-stakeholder approach

Coalition on Digital Impact (CODI) aims to bring together diverse stakeholders and coordinate existing efforts rather than duplicate work

Multi-stakeholder response needed for multifaceted issue

Summary

There is unanimous agreement that the complexity of creating a multilingual internet requires coordinated efforts across multiple stakeholder groups, with no single organization capable of addressing all necessary components alone

Topics

Interdisciplinary approaches | Capacity development | Multilingualism

Language barriers create cybersecurity vulnerabilities

Speakers

– Jennifer Chung
– Christian Daswon

Arguments

Language barriers create safety risks online as people cannot identify suspicious content or phishing attempts in unfamiliar languages

Cybersecurity is one of the important reasons why focusing on digital inclusion is so important right now

Summary

Both speakers recognize that language barriers make users more vulnerable to cybersecurity threats because they cannot identify suspicious content in languages they don’t understand

Topics

Cybersecurity | Multilingualism | Digital access

Government policy and procurement can drive multilingual adoption

Speakers

– Manal Ismail
– Jennifer Chung
– Theresa Swinehart

Arguments

Governments should integrate multilingual support into digital accessibility laws, procurement requirements, and national digital transformation strategies

De-incentivizing non-compliance through procurement processes and requiring Universal Acceptance readiness reporting in tenders

Soft enforcement mechanisms include education system requirements, government agency leadership by example, and procurement policy preferences

Summary

All speakers agree that government policy, particularly through procurement processes and regulatory frameworks, can create effective incentives for multilingual internet adoption

Topics

Legal and regulatory | Data governance | Digital access

Similar viewpoints

Meaningful connectivity requires more than basic internet access – it must enable full participation in digital services including healthcare, education, and government services in users’ native languages

Speakers

– Sophie Mitchell
– Jennifer Chung
– Manal Ismail

Arguments

Digital inclusion gaps exist not just in access but in meaningful connectivity, with people using internet only for basic functions like social media rather than essential services

Capacity building and digital skills training in native languages increases economic participation and civic engagement

Limited multilingual content availability, poor machine translations, and cultural irrelevance create vicious cycles where younger generations lose fluency in ancestral languages

Topics

Digital access | Capacity development | Multilingualism

Technical implementation challenges persist even when basic language encoding is available, requiring proper tool selection and open data availability

Speakers

– Toral Cowieson
– Audience

Arguments

Success depends on choosing development tools that have Unicode built into their technical stack, as demonstrated by differences between platforms

Font rendering and screen reader accessibility remain practical obstacles for website creators wanting to implement their native languages

Open licensing of quality text and data is crucial for AI and language models to learn multilingual approaches effectively

Topics

Digital standards | Technical Solutions and Infrastructure Development | Multilingualism

Language preservation is critical for maintaining cultural heritage and preventing massive loss of human knowledge and identity

Speakers

– Jennifer Chung
– Sophie Mitchell
– Manal Ismail

Arguments

90% of world’s languages might become extinct by end of century, representing massive loss of human culture and knowledge

Australia has lost about half of its 250 Indigenous language groups due to colonization, with many remaining languages endangered and originally oral rather than written

Limited multilingual content availability, poor machine translations, and cultural irrelevance create vicious cycles where younger generations lose fluency in ancestral languages

Topics

Cultural diversity | Multilingualism | Human rights principles

Unexpected consensus

Economic benefits of multilingual internet access

Speakers

– Theresa Swinehart
– Sophie Mitchell
– Manal Ismail

Arguments

Multilingual internet access creates economic opportunities by connecting businesses with previously unreachable consumer populations

Capacity building and digital skills training in native languages increases economic participation and civic engagement

Incentive-based approaches like tax breaks, grants, and public-private partnerships can encourage private sector adoption of multilingual technologies

Explanation

While the discussion was primarily framed around cultural preservation and digital rights, there was unexpected strong consensus on the business case and economic benefits of multilingual internet access, suggesting market-driven solutions alongside policy approaches

Topics

Economic | Digital business models | Multilingualism

AI and technology as both opportunity and threat for language diversity

Speakers

– Sophie Mitchell
– Audience
– Christian Daswon

Arguments

Question whether AI translation will accelerate multilingual internet or be part of the problem

Northern Sami Wikipedia seeing 88% bot/AI traffic vs 12% human traffic, with AI desperately trying to learn the language

In the era of advancing AI and LLMs, there are data-rich languages and many data-poor languages creating new challenges

Explanation

There was unexpected consensus that AI presents both opportunities and risks for language diversity – while AI systems are actively seeking to learn minority languages, they may also create new forms of digital divide between data-rich and data-poor languages

Topics

Digital standards | Multilingualism | Technical Solutions and Infrastructure Development

Overall assessment

Summary

The speakers demonstrated remarkably high consensus across multiple dimensions: the need for paradigm shift from English-first to multilingual-by-design thinking, the essential role of multi-stakeholder collaboration, the importance of government policy in driving adoption, and the recognition that language barriers create both cultural and security vulnerabilities. There was also unexpected agreement on economic benefits and the dual nature of AI as both opportunity and threat for language diversity.

Consensus level

Very high consensus with strong alignment on fundamental principles, implementation strategies, and the urgency of action. The implications are significant as this level of agreement among diverse stakeholders (technical, policy, civil society, government) suggests strong potential for coordinated action and indicates that the multilingual internet movement has moved beyond awareness-building to practical implementation planning. The consensus also validates the approach of treating language inclusion as core digital infrastructure rather than optional feature.

Different viewpoints

Scope of organizational mission regarding oral-only languages

Speakers

– Toral Cowieson
– Sophie Mitchell

Arguments

Unicode Consortium focuses on written languages, leaving gaps in supporting oral-only languages that lack written representation

Australia has lost about half of its 250 Indigenous language groups due to colonization, with many remaining languages endangered and originally oral rather than written

Summary

Toral explicitly stated that Unicode’s mission is limited to written languages and cannot address oral-only languages, while Sophie discussed the need to preserve oral Indigenous languages. This represents a gap in coverage rather than disagreement on approach.

Topics

Digital standards | Multilingualism | Cultural diversity

Unexpected differences

Role of AI translation in multilingual internet development

Speakers

– Sophie Mitchell

Arguments

Digital inclusion gaps exist not just in access but in meaningful connectivity, with people using internet only for basic functions like social media rather than essential services

Explanation

Sophie raised uncertainty about whether AI translation would help solve multilingual internet challenges or potentially reduce the need for true multilingual internet development. This was an unexpected point of ambivalence in an otherwise unified discussion, though other speakers didn’t directly address this concern.

Topics

Digital standards | Multilingualism | Digital access

Overall assessment

Summary

The discussion showed remarkable consensus among speakers on both the problems (language barriers to digital inclusion) and solutions (multilingual-by-design approaches, government incentives, stakeholder coordination). The few disagreements were primarily about organizational scope limitations rather than fundamental approaches.

Disagreement level

Very low disagreement level. This consensus suggests strong alignment in the multilingual internet community but may also indicate potential groupthink or lack of diverse perspectives on implementation challenges. The unified approach could facilitate coordinated action but might benefit from more critical examination of different strategies and their trade-offs.

Partial agreements

Similar viewpoints

Meaningful connectivity requires more than basic internet access – it must enable full participation in digital services including healthcare, education, and government services in users’ native languages

Speakers

– Sophie Mitchell
– Jennifer Chung
– Manal Ismail

Arguments

Digital inclusion gaps exist not just in access but in meaningful connectivity, with people using internet only for basic functions like social media rather than essential services

Capacity building and digital skills training in native languages increases economic participation and civic engagement

Limited multilingual content availability, poor machine translations, and cultural irrelevance create vicious cycles where younger generations lose fluency in ancestral languages

Topics

Digital access | Capacity development | Multilingualism

Technical implementation challenges persist even when basic language encoding is available, requiring proper tool selection and open data availability

Speakers

– Toral Cowieson
– Audience

Arguments

Success depends on choosing development tools that have Unicode built into their technical stack, as demonstrated by differences between platforms

Font rendering and screen reader accessibility remain practical obstacles for website creators wanting to implement their native languages

Open licensing of quality text and data is crucial for AI and language models to learn multilingual approaches effectively

Topics

Digital standards | Technical Solutions and Infrastructure Development | Multilingualism

Language preservation is critical for maintaining cultural heritage and preventing massive loss of human knowledge and identity

Speakers

– Jennifer Chung
– Sophie Mitchell
– Manal Ismail

Arguments

90% of world’s languages might become extinct by end of century, representing massive loss of human culture and knowledge

Australia has lost about half of its 250 Indigenous language groups due to colonization, with many remaining languages endangered and originally oral rather than written

Limited multilingual content availability, poor machine translations, and cultural irrelevance create vicious cycles where younger generations lose fluency in ancestral languages

Topics

Cultural diversity | Multilingualism | Human rights principles

Key takeaways

Language inclusion must be treated as a foundational pillar of digital infrastructure, not an afterthought or localization add-on

A paradigm shift from ‘English first’ to ‘multilingual first’ thinking is essential, with Universal Acceptance designed in from the beginning

Technology should serve languages rather than forcing languages to adapt to technological limitations

No single organization can solve multilingual internet challenges alone – success requires coordinated multi-stakeholder collaboration

Digital inclusion through language accessibility creates both social benefits (cultural preservation, safety, civic participation) and economic opportunities (reaching new consumer populations)

The underlying technology for multilingual internet largely exists, but awareness, implementation, and coordination are the primary barriers

Language barriers create safety risks online as users cannot identify suspicious content in unfamiliar languages

90% of world’s languages might become extinct by end of century without digital preservation efforts

Meaningful connectivity requires more than just internet access – it needs language accessibility, digital skills, and relevant content

Resolutions and action items

Individual participants committed to contacting two people in their networks to raise awareness about multilingual internet opportunities

Coalition on Digital Impact (CODI) established as coordinating mechanism for stakeholders working on language inclusion (accessible at codi.global)

CODI to conduct surveys of last-mile connectivity providers to identify what tools and support they need for language inclusion

CODI to gather data showing that language-inclusive tools increase internet usage to provide business case for implementation

ICANN to continue Universal Acceptance education through university modules, hackathons, and partnerships

Participants encouraged to integrate language awareness into existing educational curricula and professional programs

Government agencies and organizations urged to lead by example by implementing multilingual technologies in their own operations

Stakeholders to treat language inclusion as requirement in procurement processes and tender evaluations

Unresolved issues

How to effectively support oral-only languages that lack written representation in digital systems

Whether AI translation will accelerate multilingual internet adoption or become part of the problem by reducing incentives for native language development

How to balance open licensing needs for AI language model training with language communities’ concerns about giving up ownership rights

Specific mechanisms for quantifying and demonstrating the economic benefits of multilingual internet investment

How to address font rendering and accessibility tool limitations that prevent website creators from implementing native languages

Strategies for overcoming the vicious cycle where lack of demand reduces supply incentives and vice versa

How to scale successful local initiatives to global implementation

Methods for ensuring cultural context and localization beyond literal translation

Suggested compromises

Soft enforcement mechanisms for Universal Acceptance readiness through education, procurement preferences, and leading by example rather than regulatory mandates

Phased approach to procurement requirements – first prioritizing contractors with multilingual capabilities, then eventually making it mandatory

Working simultaneously on supply and demand sides – creating user awareness while building technical capabilities

Integrating language inclusion modules into existing educational programs rather than creating entirely new curricula

Focusing on strengthening and coordinating existing initiatives rather than launching new competing programs

Using interpretation services more effectively in international meetings by encouraging speakers to use their native languages when interpretation is available

Balancing open licensing needs with language community concerns by clearly distinguishing between granting usage rights and retaining ownership

Digital inclusion starts with Unicode, starts with the work of Unicode, which includes character encoding… for those of us in majority languages there are many things that we can take for granted. So for example, can I exchange currency, date and time format, time zone adjustments, usage of numerical symbols or decimals and commas separators used appropriately.

Speaker

Toral Cowieson

Reason

This comment was insightful because it revealed the invisible infrastructure of digital privilege – how speakers of majority languages unconsciously benefit from systems designed around their linguistic needs. It shifted the conversation from high-level policy discussions to concrete, granular examples of digital exclusion that most people never consider.

Impact

This comment grounded the entire discussion in practical reality and helped other panelists frame their responses around specific, tangible barriers rather than abstract concepts. It established a foundation for understanding how deep linguistic bias runs in digital systems.

I think there was a statistic somewhere that said, you know, 90% of the world’s languages might become extinct at the end of this century. That’s like, that is a big loss, not in terms of how we communicate, but really, it’s human culture.

Speaker

Jennifer Chung

Reason

This comment was profound because it reframed the entire discussion from a technical/accessibility issue to a cultural preservation crisis. It elevated the stakes from convenience to cultural survival, adding urgency and moral weight to the technical solutions being discussed.

Impact

This shifted the conversation’s tone from problem-solving to crisis response, and subsequent speakers began incorporating cultural preservation arguments alongside technical and economic ones. It helped establish multilingual internet access as a human rights issue.

The one thing that will accelerate it the most is actually a paradigm shift to thinking about looking at it as multilingual first as opposed to English first. To look at it by, you know, multilingual by design, universal acceptance, UA by design.

Speaker

Jennifer Chung

Reason

This comment introduced a fundamental conceptual framework that became the central theme of the discussion. Rather than treating multilingualism as an add-on feature, she proposed making it the default design principle – a complete inversion of current practice.

Impact

This concept of ‘multilingual by design’ was immediately adopted and repeated by multiple other panelists throughout the remainder of the discussion. It became the organizing principle around which other speakers structured their recommendations and policy proposals.

So instead of making languages change technology, we need to start making technology serve languages.

Speaker

Manal Ismail

Reason

This was a brilliantly concise articulation of the power dynamic at the heart of the digital divide. It captured how current systems force linguistic communities to adapt to technology rather than technology adapting to serve diverse communities.

Impact

This phrase became a rallying cry that Ram Mohan immediately highlighted as ‘powerful’ and used to transition the discussion toward practical solutions. It provided a memorable framework that other speakers referenced in their closing statements.

Understanding that having this [multilingual internet] actually makes for a safer Internet is something that we need to think about… for someone who absolutely don’t understand these characters at all, that could be a huge mystery.

Speaker

Jennifer Chung

Reason

This comment introduced an entirely new dimension to the discussion – cybersecurity. It was insightful because it connected linguistic inclusion to digital safety, showing how language barriers make people more vulnerable to online threats and scams.

Impact

This cybersecurity angle was immediately picked up by Christian Dawson, who called it ‘very important’ and incorporated it into CODI’s multi-faceted approach. It added a compelling business case and safety argument to what had been primarily framed as an equity issue.

We are fighting inertia… in the era of advancing AI, when people are using the internet directly with LLMs. There are these data-rich languages and many, many, many data-poor languages.

Speaker

Christian Dawson

Reason

This comment was thought-provoking because it identified AI as both an opportunity and a threat to linguistic diversity. It highlighted how current AI development could entrench existing language hierarchies by amplifying data-rich languages while marginalizing data-poor ones.

Impact

This AI framing added urgency to the discussion and was later reinforced by Elisabeth Carrera’s observation about LLMs ‘desperately trying to learn Northern Sami.’ It helped position multilingual internet development as a race against AI-driven linguistic homogenization.

We behave in a way that I would call the syndrome of animal farm, in the sense that we all say, okay, all languages are equal, but then in practice we behave as if we have one language that is more equal than the others.

Speaker

Roberto Gaetano

Reason

This comment was particularly insightful because it called out the hypocrisy within the very community discussing linguistic inclusion. Using the Animal Farm metaphor, he pointed out how even multilingual advocates default to English-dominant behaviors in practice.

Impact

This comment created a moment of uncomfortable self-reflection that was evident in Sophie Mitchell’s response, where she admitted to ‘ruminating’ on his point. It challenged the audience to examine their own practices, not just advocate for policy changes.

Overall assessment

These key comments fundamentally shaped the discussion by elevating it from a technical problem-solving session to a comprehensive examination of digital power structures, cultural preservation, and systemic change. The conversation evolved through several distinct phases: first establishing the invisible nature of digital linguistic privilege, then reframing the issue as cultural crisis, introducing the ‘multilingual by design’ paradigm, connecting it to cybersecurity concerns, and finally challenging participants to examine their own complicity in linguistic hierarchies. The most impactful comments didn’t just add information – they shifted perspectives, introduced new frameworks, and created moments of recognition that changed how subsequent speakers approached the topic. The discussion’s trajectory moved from identifying problems to proposing systemic solutions to confronting uncomfortable truths about current practice, creating a more nuanced and actionable understanding of digital linguistic inclusion.

How can we gather data that shows if people build tools for language inclusion at the last mile, recipients will use the Internet more?

Speaker

Christian Daswon

Explanation

Hard data is needed to convince businesses to commit time and resources to language inclusion tools, as ‘quite obvious’ isn’t sufficient for business investment decisions

How are others addressing oral languages that have no written representation from a technical perspective?

Speaker

Toral Cowieson

Explanation

Unicode’s work focuses on written languages, so there’s a gap in addressing oral-only languages that needs exploration by other organizations

Will AI translation accelerate multilingual internet progress or be part of the problem?

Speaker

Sophie Mitchell

Explanation

There’s uncertainty about whether AI translation tools will reduce the need for native multilingual internet development or actually help solve accessibility issues

How can we directly survey stakeholders connecting the unconnected to understand what tools they need for language inclusion?

Speaker

Christian Daswon

Explanation

Direct information from people doing last-mile connectivity work is needed to understand what’s missing and how to collectively build necessary tools

How can we strengthen coordination across stakeholders and clarify roles and responsibilities for multilingual internet development?

Speaker

Toral Cowieson

Explanation

There’s too much fragmented work happening in isolation, and language communities need clearer pathways to engage and contribute to solutions

How can we ensure underlying web development tools (fonts, screen readers, etc.) support diverse languages for non-technical users?

Speaker

Gabriel (audience member)

Explanation

Technical barriers prevent educators and bloggers from creating content in their native languages without deep technical knowledge

How can we make quality text and data available under open licenses for small languages?

Speaker

Elisabeth Carrera (Wikimedia Norway)

Explanation

AI systems need access to quality multilingual data, but much existing content in minority languages isn’t available under open licenses for training purposes

What are effective soft enforcement mechanisms for accelerating IDN compliance and UA readiness beyond law enforcement?

Speaker

Nicholas Fumarelli (online participant)

Explanation

Countries need alternatives to legal mandates to encourage adoption of internationalized domain names and universal acceptance standards

How can we quantify the benefits of multilingual internet access to drive unified stakeholder action?

Speaker

Sophie Mitchell

Explanation

Better data on the concrete benefits of language inclusion could help unify different stakeholders around common goals and justify investments