WSIS Action Line C8: Multilingualism in the Digital Age: Inclusive Strategies for a People-Centered Information Society

Summary

This UNESCO session focused on multilingualism in the digital age and strategies for promoting cultural diversity in creating a people-centered information society. Dr. Tawfik Jelassi opened by highlighting the stark digital language divide, noting that while over 8,000 languages exist globally, fewer than 120 are represented online, with 91% of content available in just 14 languages. This forces nearly a billion people to learn additional languages to access online resources, effectively marginalizing thousands of linguistic communities and their cultural identities.

The discussion emphasized that achieving true digital multilingualism requires systemic change beyond just technological tools. Guilherme Canela outlined UNESCO’s comprehensive approach through standard-setting, capacity building, serving as a laboratory of ideas, acting as a clearinghouse, and promoting international cooperation. He highlighted the 2003 UNESCO recommendation on multilingualism in cyberspace and the ongoing UN International Decade of Indigenous Languages (2022-2032), which has already resulted in 15 countries developing national action plans for indigenous languages.

Valts Ernstreits shared insights from working with the endangered Livonian language, emphasizing the importance of community-driven approaches and speaker involvement programs. He stressed that technology development must be community-specific and defined by actual community needs rather than imposed solutions. Sofiya Zahova presented successful examples from Nordic and Arctic regions, including Iceland’s initiative to maintain its language presence in AI models and Sámi language technology projects led by indigenous institutions.

David Waweru focused on Africa’s challenges and opportunities, noting that despite Africa being home to over 2,000 languages, fewer than 0.1% of websites contain African language content. He highlighted promising initiatives like the African Storybook project and various digital publishing platforms that are beginning to bridge this gap. Elena Plexida addressed universal acceptance challenges, revealing that only 26% of email servers support internationalized email addresses, making digital inclusion a matter of linguistic justice.

The session concluded with calls for multi-stakeholder collaboration involving governments, communities, private sector, and civil society to ensure no language or culture is left behind in the digital future.

Keypoints

## Major Discussion Points:

– **Digital linguistic divide and underrepresentation**: Only 120 of the world’s 8,000+ languages are represented online, with over 91% of content available in just 14 languages, creating barriers for nearly a billion people who must learn additional languages to access online resources.

– **Community-driven approaches to language preservation**: Successful examples include hiring native speakers to develop digital resources (Livonian Institute), community-led technology development, and indigenous language initiatives in Nordic/Arctic regions that prioritize cultural sensitivity and self-determination.

– **Universal Acceptance technical challenges**: Only 26% of email servers support internationalized email addresses, and most websites reject non-Latin script email addresses, creating digital identity barriers for users of languages like Arabic, Thai, Greek, and others.

– **Policy frameworks and multi-stakeholder collaboration**: The need for systematic change through UNESCO’s 2003 recommendation on multilingualism, the UN International Decade of Indigenous Languages (2022-2032), and coordinated efforts between governments, communities, private sector, and civil society.

– **Cultural and creative industries in digital spaces**: Examples from Africa showing how digital platforms can amplify indigenous storytelling, oral traditions, and local publishing, while addressing the dominance of colonial languages and promoting mother-tongue literacy.

## Overall Purpose:

This UNESCO-led WSIS Action Line C8 session aimed to address multilingualism in the digital age and develop inclusive strategies for cultural diversity in building a people-centered information society. The discussion focused on identifying barriers, sharing successful approaches, and coordinating multi-stakeholder efforts to ensure linguistic diversity is preserved and promoted in digital spaces.

## Overall Tone:

The discussion maintained a professional, collaborative, and solution-oriented tone throughout. While speakers acknowledged serious challenges and expressed concern about the digital linguistic divide, the overall atmosphere was constructive and forward-looking. There was a sense of urgency balanced with optimism, as panelists shared concrete examples of successful initiatives and emphasized the importance of collective action. The tone remained consistently inclusive and respectful of cultural diversity, with speakers building upon each other’s insights rather than debating opposing viewpoints.

Speakers

**Speakers from the provided list:**

– **Davide Storti** – UNESCO, Paris (Session moderator)

– **Tawfik Jelassi** – Assistant Director General for UNESCO Communication and Information

– **Guilherme Canela de Souza Godoi** – Director for Digital Inclusion and Policies and Digital Transformation at UNESCO

– **Valts Ernstreits** – Director and Representative of Latvia and University of Latvia in Livonian Institute

– **Sofiya Zahova** – Director of VG’s International Center for Multilingualism and Intercultural Understanding from Iceland (joined online)

– **David Waweru** – Publisher and Researcher (joined online)

– **Elena Plexida** – Vice President for Government and IGO Engagement at the Internet Corporation for Assigned Names and Numbers (ICANN)

**Additional speakers:**

None – all speakers mentioned in the transcript were included in the provided speakers names list.

# UNESCO Session on Multilingualism in the Digital Age: Discussion Report

## Executive Summary

This UNESCO-led WSIS Action Line C8 session addressed multilingualism in the digital age, focusing on strategies for promoting cultural diversity in creating a people-centred information society. Moderated by Davide Storti from UNESCO Paris, the discussion brought together international experts to examine the digital language divide and explore solutions through policy frameworks, community-driven approaches, and multi-stakeholder collaboration.

The session highlighted stark statistics about digital linguistic exclusion: while more than 8,000 languages exist globally, fewer than 120 are represented online, forcing nearly a billion people to learn additional languages to access online resources and effectively marginalising thousands of linguistic communities.

## Opening Remarks and Problem Definition

Dr Tawfik Jelassi, Assistant Director General for UNESCO Communication and Information, opened the session with a powerful metaphor describing the digital world as “a vast library” where “if only a few languages are represented by books on the shelves of that library, think how many civilisations, how many histories would be left unheard and unread.”

Jelassi quoted UNESCO Director General Audrey Azoulay: “languages are more than tools for communication. Languages are the very condition of our humanity.” He emphasised that achieving true digital multilingualism requires systemic change beyond technological tools, highlighting UNESCO’s role in developing a global roadmap for language technologies through consultations running until the end of the week.

## UNESCO’s Policy Framework and Initiatives

Guilherme Canela de Souza Godoi, Director for Digital Inclusion and Policies and Digital Transformation at UNESCO, outlined the organisation’s approach to digital multilingualism. He referenced UNESCO’s 80th anniversary and its constitutional mandate regarding the “free flow of ideas and information,” connecting this to current multilingualism challenges.

Canela highlighted the 2003 UNESCO recommendation on multilingualism in cyberspace, which requires government reporting on policy implementation. However, he revealed that only 60% of reporting countries are actually implementing policies in this area, while many governments fail to report altogether.

He emphasised the ongoing UN International Decade of Indigenous Languages (2022-2032) as a crucial opportunity, reporting that 15 countries have developed national action plans for indigenous languages, with 20 additional countries currently developing such plans.

## Community-Driven Approaches: The Livonian Case Study

Valts Ernstreits, Director and Representative of Latvia and University of Latvia in Livonian Institute, provided insights from working with the critically endangered Livonian language, which has “less than 20 language speakers scattered all across the country.”

Ernstreits revealed a fundamental challenge: “if you do not able to produce any digital data so you cannot have any digital technology at all and this is true for quite many world’s languages who cannot simply enter digital domain.” He emphasised the importance of speaker involvement programmes, where native speakers are hired to develop digital resources, creating both technological advancement and language revitalisation.

The Livonian case illustrated challenges faced by endangered languages in the AI age, where limited data availability makes it extremely difficult to create AI solutions or large language models, potentially widening the digital divide between well-resourced and endangered languages.

## Nordic and Arctic Success Stories

Sofiya Zahova, Director of VG’s International Center for Multilingualism and Intercultural Understanding from Iceland, presented three specific examples of successful initiatives:

1. Iceland’s government investment in language technology, demonstrating how low-resource languages can assert their digital presence through strategic policy intervention

2. Sámi language technology projects led by indigenous institutions themselves

3. Roma digital engagement initiatives

Zahova warned against “extractive digitisation,” emphasising that “we must avoid the trap of extractive digitisation and ensure that communities, especially indigenous and minority ones, are the authors of their digital futures.” She stressed that multilingual accessibility should be treated as critical infrastructure similar to healthcare, requiring state support and systematic investment.

## African Perspectives and Digital Transformation

David Waweru, Publisher and Researcher joining online, focused on Africa’s unique challenges in digital multilingualism. He presented statistics revealing that despite Africa being home to over 2,000 languages, very little web content exists in African languages, with colonial languages dominating digital spaces.

Waweru highlighted several promising initiatives bridging the digital language gap:

– World Reader’s mobile library providing access to books in local languages

– The African Storybook project offering culturally relevant content in multiple African languages

– Various digital publishing platforms amplifying indigenous storytelling and oral traditions

He referenced the African Union’s Smart Africa Coalition, which launched an AI for Africa Blueprint involving 40 nations, representing a coordinated continental effort to address digital language divides through regional cooperation.

## Technical Infrastructure and Universal Acceptance

Elena Plexida, Vice President for Government and IGO Engagement at ICANN, addressed universal acceptance challenges from a technical infrastructure perspective. She revealed that “only 26% of tested email servers support email addresses in internationalized scripts,” making digital identity “invisible or invalid” for users of non-Latin scripts.

Plexida described universal acceptance as “not a technical luxury, it’s a digital rights imperative,” connecting technical standards to fundamental questions of digital inclusion. She explained that most websites reject email addresses in scripts such as Arabic, Thai, and Greek, creating systematic barriers to digital participation.

She highlighted the role of government leadership in driving change through procurement policies and public service design, and mentioned that ICANN will launch new internationalised domain names in 2026, providing opportunities for more domains in local scripts.

## Key Challenges and Future Directions

The discussion identified several critical challenges requiring ongoing attention:

**Resource and Capacity Constraints**: Many endangered language communities lack the resources to enter the digital domain, requiring innovative funding and support mechanisms.

**Data Scarcity**: Limited data availability for low-resource languages creates barriers to developing AI solutions and language technologies.

**Implementation Gaps**: Significant gaps exist between policy development and actual implementation of multilingual digital initiatives.

**Technical Barriers**: Universal acceptance standards require consistent implementation across millions of existing applications and websites.

**Future Action Items** identified include:

– Continuing UNESCO’s consultation process for the global roadmap on language technologies

– Improving government reporting on the 2003 UNESCO recommendation implementation

– Leveraging the UN International Decade of Indigenous Languages more effectively

– Preparing for ICANN’s 2026 launch of new internationalised domain names

– Strengthening multi-stakeholder collaboration across sectors

## Conclusion

The session demonstrated that multilingualism in digital spaces represents both a technical challenge and a fundamental question of cultural survival and digital rights. The discussion revealed both the urgency of the digital language divide and the potential for coordinated action to address it through community-led approaches, policy intervention, and technical innovation.

Jelassi concluded with an interesting observation about the session’s attendance: “I see 20 women and less than 10 men,” raising questions about demographic engagement in multilingualism discussions. Moderator Storti encouraged participants to contribute to the WSIS+20 review process, emphasising the importance of continued engagement in addressing these critical digital inclusion challenges.

The session highlighted that successful digital multilingualism requires ensuring that technological solutions serve linguistic communities rather than being imposed upon them, with community agency and self-determination emerging as crucial principles for future work.

Davide Storti: in 30 seconds. Thank you. Hello, good morning, everyone. I’d like to start the session. This is a session on action C8. This is Davide Storti from UNESCO, Paris. The title of the session is Multilingualism in the Digital Age, Inclusive Strategies on Cultural Diversity for a People-Centered Information Society. We have here representatives from different sectors joining us as panelists. But before all, I think I will leave the floor to and Dr. Tawfik Jelassi, Assistant Director General for UNESCO Communication and Information, for giving the initial remarks to this session.

Tawfik Jelassi: Thank you, Davide. Ladies and gentlemen, colleagues, good morning to all of you and thank you for joining this session on OASIS Action Line 8. UNESCO has been in charge of six action lines out of 11, and this is a very important one, focusing, as you know, on multilingualism in the digital age, how to build inclusive societies, people-centered communities. And of course, yesterday we talked in the opening session about the mission of OASIS and the vision for it since 2005, which is very much about building this people-centered development-oriented information society. UNESCO has been deeply involved in this topic, mainly to preserve, revitalize and support cultural diversity and identity, while safeguarding linguistic diversity as a cornerstone of an inclusive information society. And let me here quote Madame Audrey Azoulay, the Director General of UNESCO, who once said, languages are more than tools for communication. Languages are the very condition of our humanity. The question today, what makes us human? Since the machine also today can master large language models, so what used to be the unique capability of a human’s languages to talk to each other, the machine today is starting imitating that, or mastering that. So obviously, the quote I just mentioned Mr. David reminds us that addressing the digital divide is a key priority because we don’t want lesser heard voices to be silenced and we want to protect the world’s cultural and linguistic heritage. However, despite the fact of recognizing language as a fundamental human capability, our digital infrastructure is far from being inclusive. Let me make a quick metaphor. Think of the digital world as a vast library. If only a few languages are represented by books on the shelves of that library, think how many civilizations, how many histories would be left unheard and unread. We must ensure that this library reflects the full spectrum of human experience, where every language has a place, every voice can be heard, and every culture can contribute. Therefore, a major challenge in building an inclusive digital ecosystem is the persistent lack of multilingualism in cyberspace. Our study has revealed that in the world there are more than 8,000 languages. How many of them are online? Not more than 120. Nearly a billion people must master additional languages in order to access online resources, and this clearly reveals a system that is overwhelmingly favoring a handful of dominant languages. In fact, while 5.5 billion people are connected to the internet, over 91% of all online content is available in just 14 languages. and many other languages, leaving thousands of other languages invisible online. This is not only restricting communication and access to information and knowledge, this marginalises personal and cultural identities. So this is the challenge we are facing and obviously we would like sophisticated technologies like AI and generative artificial intelligence can help bridge this linguistic gap or linguistic divide that we have been facing. Our commitment at UNESCO is reflected by our 2003 recommendation, which focuses on the promotion and use of multilingualism and universal access to cyberspace. And these are of course measures that promote local content creation and provide equitable access in all languages. UNESCO continues to lead these principles through WSIS Action Line 8. Furthermore, UNESCO has recently kicked off the United Nations International Decade of Indigenous Languages. The start was 2022 and this decade will run until 2032. And this obviously underscores the urgent need to preserve, revitalise and promote indigenous languages, especially in cyberspace. This can only be achieved through collaboration with governments, communities and stakeholders. Looking forward, UNESCO is also developing a draft roadmap for language technologies and multilingualism. This framework is currently open for consultation until the end of this week. Its aim is to harness… AI and digital tools to build inclusive, equitable digital ecosystems that promote linguistic diversity and uphold multilingualism. I invite you all to engage in this process and this open global consultation. Let me conclude by saying that only by working together, governments, communities, private sector, civil society, technical community, only by working together can we ensure that no language and no culture is left behind in our digital future. We must call for ethical data governance, inclusive digital policies, and equitable access to language technologies. The lack of resources for underrepresented communities coupled with mainstream use of handful languages is a challenge that you must work together to address. Thank you.

Davide Storti: Thank you very much, Dr. Gelassi. And now we’d like to introduce the panelists which are guiding our discussion today. We have Mr. Guilherme Canela, Director for Digital Inclusion and Policies and Digital Transformation at UNESCO. We have Mr. Valts Ernštreits, the Director and Representative of Latvia and University of Latvia in Livonian Institute. We have online, joining us, Sofiya Zahov, Director of VG’s International Center for Multilingualism and Intercultural Understanding from Iceland. We also have online Mr. David Waweru, Publisher and Researcher. And we have, if you’re to my right, Madam Elena Plexida, Vice President for Government and IGO Engagement at the Internet Corporation for Assigned Names and Numbers, ICANN. So, I’d like… I would like to start with the first intervention I would like to ask Mr. Canela. Realizing the true multilingualism in digital space and age requires more than just tools, it demands systemic change. From your perspective, what are the most strategic levers, policies, partnerships or incentives that can accelerate the integration of linguistic diversity into the global digital ecosystem, particularly for underrepresented communities? Could you give us some perspectives of your own?

Guilherme Canela de Souza Godoi: Thank you everyone. Good morning. I notice in a very glad manner that this issue is gaining lots of momentum because in the last few weeks I have seen Valts more than my children. So it seems that we are actually discussing a lot this issue in different forums, which is a good thing, don’t you agree Valts? Although we need to stop that. I will actually go a bit further on different issues that ADG Gelassi already started mentioning in his introductory remarks to just show some extra elements of these policies and multilateral discussions that UNESCO has been leading from 80 years now this year. We are also, as the entire UN system, we are celebrating our 80th anniversary. And in the very first paragraph of the UNESCO constitution, we have there that one of our areas of mandate is to promote and protect the free flow of ideas and information and so on. As you can imagine, to actually fulfill this ultimate goal, We do need to have multilingualism. There is no free flow of information, only with, as A.D. Djilas put, a handful of languages being represented. So, to respond to the first part of your broader question, our challenge is how, as the leading UN agency in charge of protecting and promoting freedom of expression, so the Article 19 of the Universal Declaration, the right to impart, to speak, but also the right to seek and receive. So, when we are doing policy towards the protection and promotion of this very complex right that is the right of freedom of expression, that’s not only about speaking, it’s also about accessing information and knowledge and ideas. So, how multilingualism becomes a cross-cutting element in the overall tools that are protecting this right. So, if you look into the history of UNESCO in terms of the guidelines for broadcasting regulation, which I’m seeing from previous years, not even talking about the internet, the issue of multilingualism in this media ecosystem was already there. So, this is important, how we do it, how we create this using the five areas that UNESCO normally does when we are doing international assistance and cooperation. So, the first is a standard setting, and A.D. Djilas already mentioned, for example, the 2003 recommendation. So, this is a standard setting, right, saying governments, you need to do A, B, and C, and you need to report back, and I will come to a second, to that recommendation. The second thing is capacity building, and we have been doing that a lot during the last many years, and one essential element of the decade of indigenous languages that A.D. Djilas already mentioned, and I will speak about that a bit more in details, all the universal acceptance that we are doing with ICANN, it’s a lot about capacity building. The third element is being in a laboratory of ideas and the global roadmap that I’m sure Valts will mention or that was born also through a kind of an interesting breakthrough conference since UNESCO further discussing cutting edge technology and multilingualism is about that, how we think about the future, right? How we serve in the UN system as UNESCO as in a laboratory of ideas. The other element is being a clearing house. So how we put together these different knowledge. So if you look into the website of the International Decade of Indigenous Languages, you will see a lot of content there that’s not generated by UNESCO, but we try to be an aggregator of the conversation around in this case, indigenous languages on this. And finally, we should be a promoter of international cooperation. That is partially what we are doing here, what we are doing with the decade, what we are doing in our association with ICANN or with the VG Center in Iceland and so on. So I don’t want to take much time off the speakers just to go deeper into two or three of these points for you to have an idea. So ADG already mentioned the 2003 recommendation. So you see in the very first year of the WSIS Summit in Geneva agenda, UNESCO was already approving a global standard setting document to promote multilingualism in the cyberspace. You can think, well, this kind of title is a bit outdated today, but if you look into the key recommendations that are in that document, and then you adapt to the new technologies, these still very much valid. And although we actually had many achievements in- In promoting multilingualism in the cyberspace, there is still a lot to be done. So for you to have an idea, although this is a recommendation that requires member states to report back to us every four years, we are still far from an ideal process of reporting. Several of the governments, they don’t report back on what they are doing. And among those that report back, only 60% reported that they are actually implementing policies on this area. So first call for action here. Next year we will start a new cycle. Please help us to stimulate your governments to report back, because the first step to improve any policy is that we need to know what is happening. We need to know what they have in mind, what are the challenges, what are the needs. During the decade of indigenous languages, also already mentioned by ADG Jelassi, let me just give you two or three important facts there. The decade started, 2022 was approved, but actually started in 2023. And right now we already have 15 countries with national action plans on indigenous languages. 20 are discussing their plans in this very moment, and we are providing technical assistance. We are producing with the Global Task Force some guidelines for the governments to do this in a truly inclusive way. The Global Task Force has four important ad hoc groups, one on education, another in languages transmission, another in policies, and another in digital. So please engage with that. Currently, actually, there is a survey open until this weekend about this issue to try to understand what’s going on with indigenous languages worldwide. So this is an important caucus. We need to take advantage of this window of opportunity with the decade that ends in 2022. to really move the needle in this area of multilingualism technology and Indigenous languages in particular. Then I want to, the thunder of ICANN, but I just want to say that we are very happy with this partnership for universal acceptance. I have some data here that my team put for me, but I will leave to you to share the good news on that. And we think this is not one, as I said, it’s not one thing or the other. We need to have a collective action here with many angles, and universal acceptance is one of these angles, is a very important one, and we need to move ahead with that. And then on this idea of not only having events about these issues, like this one, like the one we had in February, like the IGF, we decided to congregate several of these inputs in a global roadmap for languages and technologies. And we had a drafting committee who did the first proposal. This proposal is open, as Adijit Jalasi mentioned, for consultations until the 12th of July. And we hope this roadmap would be an interesting document for helping us in moving to the next steps. And then you were going to mention there is also an action plan, I guess, there is the next step and so on. So David, this is a bit where we are with these different movements. But of course, glad to take questions later on. Thank you.

Davide Storti: Thank you very much, Guilherme. And now I’d like to move immediately to Dr. Valts Ernstreits. So we heard about the measuring progress is indeed one of the challenges, maybe ahead of the WSIS plus 20 review, and maybe we can have some insights, particularly on the, you know, the capacity and resource challenges of many endangered languages. So what promising community-driven approaches or technologies, you know, crowdsourcing lexicons Open Source, NLPs. What kind of approaches have you been seeing succeeding revitalizing languages like Livonian and how can this be adapted elsewhere?

Valts Ernstreits: Thank you for the question. So I really am in two capacities here. So one is Director of Livonian Institute, which is dealing with Latvians in indigenous Livonian language. And the other is actually I’m also a co-chair of aforementioned ad hoc group on digital equality and domains. And just for a context, Livonians are kind of a very tiny linguistic group. So we have community around but less than 20 language speakers, which are scattered all across the country. And in a sense, it represents the future of many endangered language communities due to urbanization, which the landscape is changing. And from our perspective, so when we develop technology, we kind of we try to benefit as much as possible from the digital world. But we approach technology from two kind of perspectives. So one is sustaining language. And here it’s like approach where the technology is perceived as kind of prosthetics given or enabling us to replace certain areas in the language ecosystem that are lost over the time, such as language environment or education and others. But another perspective, how we approach technology is we’re looking towards unlocking heritage and unlocking creative process as language and intangible culture, heritage and culture, they are on one hand, they’re bound together strongly. But on the other hand, Intangible heritage is locked behind the language wall simply because like in our case it can only be accessed by those community members who are proficient in the language so we look how we can exploit technology to provide access to cultural contact for those not proficient in language anymore. So what we develop is like a broad spectrum of different digital tools and resources but maybe the first point looking at that is that the technology is community-specific always so it is defined by the community needs and prioritized also by community needs and language actual situation. So for example in Livonian case if we have we have very complex grammar so one of the first things that we developed was one-click access to all the grammatical information. Another for example so you don’t have language situations when you can hear how the language sounds so we’re trying to supplement audio recordings for materials so those who want can hear how the language sounds. And why this is important is also because in this process resourcing is a key so if you look beyond top hundred of world’s languages so a lack of resources starts to become a problem more and more and more and so it’s very important to understand where does the effort goes but resource available resources also define feasibility. So basically if you do not able to produce any digital data so you cannot have any digital technology at all and this is true for quite many world’s languages who cannot simply enter digital domain or the same problem that we have in Livonian case if we have limited data it’s very hard to create AI solutions or large language models. So you know one approach that we have been implementing during the past three years, is the speaker’s involvement program. So the idea is to hire those few Livonian speakers to work for at institute. And on the task connected to the development of digital resources and tools. So they do not have linguistic skills, and they do not possess developer skills, but they have one special power, the proficiency in the language. So this is the resource that we lack very much, which we need to build the understanding for developing digital tools. So have gained an enormous resource for crucial tasks, like giving the voices, etc. So, from one perspective, this is our gain, but it also contributes back. Because it also offers them, besides digital literacy, working with those digital tools, but it also provides them with language environment. So and because of that, in last three years, we see kind of a real clear rise of real revitalization

Davide Storti: thing.

Valts Ernstreits: And another very important aspect for us is that by community being closely involved, so they might not say how to build tools or what kind of specific action should be taken, but they can definitely tell what’s missing, what’s wrong, what doesn’t work. And this is something that we gain back and which improves greatly what we do. So community involvement is actually very crucial when developing technologies. So It is, and it’s actually very beneficial as well, we have worked with developers for quite a long time and maybe two interesting approaches that I have experienced. So one is, and both could be answered with yes, but, so one approach is that, well, it’s too expensive or who needs this, which is basically, yes, but it’s like saying someone who is on life support just telling him that it’s too expensive to keep you alive, so we do deal with someone or somebody else. And another is like, well, here’s technology, please take and use it. But there is example again from Livonian, that there is a major technology technology company that made Livonian keyboard and it’s downloadable. The problem is that it’s missing a couple of characters. So it is there, but it doesn’t work. So past 10 to 20 years, we have moved, but generally if we look at the technology landscape, we are moved towards technology being able to actually suit needs of every individual. So social media was that launched it powerfully, but AI is something that has brought us closer to this than ever before. So because it is not just a technology, but it is really personally trained technology. But we cannot achieve true customization if we don’t solve language issue first, because this is crucial for everything that follows. And beyond also these are cultural context, tradition, governance, and other individualization issues. So the one way that one initiative, which is now is the roadmap of the multilingualism in digital era. And there are lots of things. and in development of the action plan so that we can reach the goal.

Davide Storti: Thank you. Thank you very much for these insights. And I would like now to pass the floor to Dr. Sofiya Zahova. You know, one of the unique values of WSIS is that it provides a multi-stakeholder space to debate about cultural diversity in the formation of society. It’s something we take maybe sometimes for granted, but it’s not the case, really. So how can efforts to promote linguistic diversity in digital spaces be more effectively linked with the protection of cultural diversity? Dr. Zahova, are there examples that have successfully supported both?

Sofiya Zahova: Thank you, Davide. I’m honored and delighted to join you today on this important panel, but even more pleased to receive a question that would allow me to reflect on rather positive examples from our work, our research. and all the forums we have organized as the national promoter of the International Decade of Indigenous Languages. And I would like to highlight three examples that are mostly from the geographic area we operate in, that’s the Nordic and Arctic countries, that demonstrate how engagement with language can become a vehicle for preserving identity, fostering inclusion and encouraging community participation. The first example I will give is the Icelandic case. You might be aware that only around 350,000 people speak our national language and Iceland aims to become an example of how low-resource language can assert a presence in the digital era. In collaboration with global technology developers such as OpenAI, investments have been made on behalf of the government in an initiative to enable Icelandic being present and respected in language technology models, applying an approach grounded in cultural sensitivity, awareness of ethnic diversity and the social role of language. The first phase of the program, that was from 2019 to 2023, was focused on data collection and production of language resources, a very important point that also Vaz has mentioned, while currently there is the second language technology program that focuses on implementation, but also on AI. In this respect, Iceland and the Icelandic Language Technology Centre are in close collaboration with UNESCO on reflecting on a global initiative that will touch upon ethical aspects and questions of bias when it comes to language technology. Technology and AI. And looking beyond the national borders, actually Iceland also supported a report that looked at languages with less than half a million speakers in the Nordic region. And in this report, there is a creation of a framework that other communities can adopt. This model suggests that even the so-called long tail or minoritized languages can become medium resourced when supported by policy, data strategy and inclusive partnerships. Then the second example I would like to reflect on is two initiatives led by indigenous people’s community in the Arctic region. Over the past two decades, Sámi language technology has seen pioneering development through projects such as Gela Tekno or Divum. These projects are not just about building keyboards and spell checkers. They were actually developed under the leadership of the Sámi institutions and the Sámi parliaments in the Nordic countries, for instance. So they actually represent a really good model of assertion of cultural rights and self-determination. Similarly, the 2022 launch of Inuktitut as a Facebook interface language that resulted from META’s partnership with Nunavut Tungavik Incorporated and the Pirurivik Center represents a breakthrough for the Inuit languages. It allowed Inuit users to navigate social media in their mother tongue, but also it involved and it’s still involving the creation of new terminology that respects cultural meaning. And then the third example I would like to reflect on is. As one of the world’s largest transnational minorities, Roma populations are often the most underprivileged linguistically, culturally and socially. Despite systematic barriers, Roma communities have shown remarkable digital engagement, particularly in the use of mobile technology and social media. And for me, as a researcher whose field is researching Romani literary heritage, it’s been really inspiring to see how uplifting it is when we put a digital old material from the 19th century that shows that Romani language was a language with written tradition. But this is an unknown fact. Now the digital technologies allow us to digitize this material, put it online, make it accessible and let the communities engage with this material. Then of course, since the last year, we have the inclusion of Lach Romani, a variant of the Romani dialect spoken globally on Google Translate. This may appear as a significant step toward linguistic visibility, but has raised numerous concerns among scholars and community members alike. The concerns were mainly focused on questions of representation, ethical considerations, community involvement and the methodological ground for selecting one dialect over others without transparent consultation or participation from the Romani communities themselves. Over the last years, as I have mentioned, we’ve been engaged as the national promoter of the International Decade of Indigenous Languages and we served as a platform for multi-stakeholder dialogue. So I would like to share with you just a few important points that have been reoccurring in our research and also during the initiatives and forums that we organized in the framework of the international decade, be them in the framework of larger forums as the Arctic Circle Assembly, held annually in Iceland, or hosted in our center. Multilingual accessibility and preserving linguistic diversity must be viewed as institutionalized and state-supported infrastructure as critical as roles of health care. This refers to something already mentioned earlier by Valts and something I’ve been hearing throughout many sessions, including at this forum. This includes investing in open data sets, community archiving, and local capacity for content creation. Efforts to promote linguistic diversity in digital space are most impactful when they are community-led, culturally and ethically grounded, and supported by long-term partnerships. We must avoid the trap of extractive digitization and ensure that communities, especially indigenous and minority ones, are the authors of their digital futures, and they decide what has to be made available online and how digital technologies have to be utilized. There is a pressing need for international cooperation and a global framework to ensure the inclusion of low-resource languages and underrepresented cultures in the design, training, and evaluation of digital technologies, particularly AI. And to conclude, the examples that I have highlighted show that when linguistic diversity is treated not as a technical challenge, but as a right and a cultural asset, digital tools can become powerful enablers of cultural expression and participation. They also demonstrate the importance of Inclusive design where language technologies are shaped with and by the communities they serve. To effectively link linguistic and cultural diversity, we must invest in infrastructures that go beyond access to ensure agency, representation and digital self-determination.

Davide Storti: Thank you very much for this intervention. I’d like to now give the floor to the next speaker. I hope we have time to give also for some questions in exchange. So I’d like to pass the floor to Dr. David Waworu, and expanding a bit the reach of this discussion and Action Line C8, in the context of rapid digital transformation, how can we ensure that digital spaces genuinely reflect and promote the full spectrum of cultural diversity, not only in terms of linguistic diversity variety, but also to the diverse cultural and creative industries, music, publishing, film, storytelling and indigenous knowledge systems. So Dr. Waworu. Thank you very much, Davide.

David Waweru: And allow me to use Africa as a case study to reflect on how to harness multilingual digital transformation to preserve cultural and linguistic heritage. And I’ll frame my reflection through six critical lenses. The first one is the challenge, and I’ll focus on the diversity in a digital age. The fact that linguistic and cultural diversity is a cornerstone of an inclusive digital society. And yet, when you look at Africa, progress has been slow, and many African languages and communities face systemic barriers. Thank you very much. So, I am very happy to be here today to talk about Africa’s digital transformation. We have seen the rapid digital transformation in the last several years online. From limited infrastructure to under-representation on major platforms. The rapid digital transformation risks leaving these voices behind, undermining linguistic rights and access to transformation. I’ll give you two examples. The first one is Africa being home to over 2,000 languages, which is about a third of the world’s languages. This represents an immense cultural wealth, however. This abundance is not reflected online. Fewer than 0.1% of websites have content in African languages. For example, more websites use Norwegian, spoken by about 4 million people, than use Swahili, which is spoken by over 200 million people. A stark illustration of the digital linguistic divide. The second example is Africa being home to over 2,000 languages, which is about a third of the world’s languages. This represents an immense cultural wealth, however. This abundance is not reflected online. Fewer than 0.1% of websites have content in African languages. For example, more websites use Norwegian, spoken by about 4 million people, than use Swahili, which is spoken by over 200 million people. A stark illustration of the digital linguistic divide. The second example is Africa’s publishing, which, according to the latest African book industry report UNESCO accounts for 5.4 % of the global market in 2023. And yet, despite this growing economic footprint, the sector still heavily relies on foreign publishing houses and imported content. Which means that African linguistic diversity and rich storytelling traditions remain underrepresented, especially in digital spaces. This limits the creation and accessibility of locally relevant, culturally rounded content. The second lens that I would like to reflect on is the cultural and creative industries. That is African voices in digital spaces. Africans’ cultural and creative industries, from music to film, are vibrant and increasingly global. And you might think of Afrobeats, for example, in music, or Nollywood films from Nigeria. And yet local languages and narratives often remain sidelined on global digital platforms, leading to a one-size-fits-all digital culture. African authors and storytellers have earned global acclaim, but historically the industry has been dominated by colonial languages, and local narratives have struggled to find visibility. Solutions for this digital media may offer new opportunities. The rise of digital publishing and indigenous language literature is reshaping this landscape. What was long dominated by English, French and Portuguese is beginning to open up, making African stories more accessible worldwide and fostering greater inclusion across the continent. For example, more books and articles are now published in Kiswahili, spoken by over 200 million people, and Yoruba, Amharic and other major languages. And these are already reaching global audiences instantly. How about indigenous storytelling and knowledge? Africa’s rich oral traditions and indigenous knowledge systems can be amplified by digital tools. When these are recorded and shared online, for example, via podcasts, videos, digital archives, they don’t just preserve, they also provide valuable perspectives on global issues. Example, community media and knowledge hubs have used local oral histories to combat misinformation and countering climate disinformation, for example, by leveraging indigenous knowledge in storytelling. Protecting and promoting such diverse content online enriches the global information. partners with schools and libraries and publishers to put thousands of books including local languages into the hands of readers. In Kenya for instance world readers mobile library and similar platforms for example local e-book stores are expanding the reach of publishers and authors by offering convenient digital access to books countrywide. This tackles the access gap and forced us a reading culture in mother tongues. How about the African Storybook? An open access initiative providing storybooks in dozens of African languages. It offers a library of culturally relevant children’s stories developed by Africans authors and educators that can read, that can be read, downloaded and even translated or adapted widely used for early literacy in multiple countries. African Storybook is supporting mother tongue education and literacy by making over a thousand storybooks available in 40 plus languages of Africa via web and a mobile app. This has become a cornerstone for early reading in local languages. How about Abound to AI? A Kenyan-based AI initiative building language technology by Africans for Africans. Abound to AI leverages advanced models such as chat GPT to develop speech and Dr. Klaus Niemann. The fourth area is Snaplify. One of Africa’s largest platforms. It provides a digital library and text book distribution system used by millions of students and thousands of schools across the continent. Other notable platforms include Nigeria’s Genti for African language audio books, Cameroon’s Ecology, how to localize AI and language tech that hold the promise to bridge linguistic divides. Second one is community innovation across Africa. Researchers and volunteers are rising to the challenge and local AI projects like Ghana’s Kahaya Fortui and Pan-African Masahane NLP community show that when communities outsource data and expertise, they can create voice recognition and translation for languages that big tech overlooks. The fifth lens is basically policy frameworks, ethics and alignment. Inclusive policy is quite important to truly democratize digital spaces. Strategic roadmaps are also critical and African leaders and institutions are actually increasingly recognizing these needs. For example, the African Union’s and Dr. David S. Nguyen. I would like to start by introducing the Smart Africa Coalition comprising 40 nations and recently launched Artificial Intelligence for Africa Blueprint. And the key is implementing these major roadmaps. And the final lens that I would like to reflect on is the need for multi-stakeholder action Again, earlier it was alluded to that no single actor can solve the digital language divide. It requires a broad coalition policy makers, publishers, booksellers, librarians, the creatives across board, authors, tech companies, local startups, academia, civil society, all needed. And of course, the need to empower local voices, global partnerships are also welcome and needed where big and small get together. Tech giants are partnering with grassroots organizations as well. And final comment is that stakeholders should prioritize linguistic diversity in their digital agendas. Thank you.

Davide Storti: Thank you very much. I would like to move further to ICANN. I did it a number of times. Dr. Elena Plexida. And I would like to ask a question about universal acceptance, which adoption remains uneven. And so can you tell us maybe how do we better raise awareness among developers, providers, policy makers, but also what are the most urgent technical regulatory changes on, for example, ensuring how email addresses and domain names of course, your domain in all scripts are fully supported across all the platforms. Thank you.

Elena Plexida: Thank you, Davide, and thank you to UNESCO and the WISH community for bringing us together today. Thank you for giving me the opportunity to draw attention to something that is both very technical and deeply human at the same time, which is how people are or are not recognized in the digital world because of the scripts and languages we use. What is universal acceptance and why it matters? Universal acceptance, or UA, is the simple but essential principle that all domain names and email addresses, regardless of script, language, length, should work in every software application, which, as you said, unfortunately is not the case. It’s a foundational requirement for a multilingual Internet. On the Internet, without universal acceptance, users that register a domain name or create an email address in their native language, say Greek in my case, or Amharic or Thai or Arabic, they will find that websites reject their email addresses as invalid, or, I’m sure you have seen that, systems process and display characters in a very weird manner, particularly when it comes to right-to-left scripts, such as the Arabic script. In human terms, what does that mean? A person’s digital identity, tied to a language or a script, is rendered invisible or invalid in that way. And that contradicts everything we stand for under WSIS and SDG 9c. It contradicts a universally accessible, inclusive, and user-centric Internet. So, universal acceptance, although it’s a very technical thing, it’s not just a technical upgrade. It’s a matter of digital equity and linguistic justice. And I will quote Dr. Zelassi, who has said, universal acceptance is not the technical luxury, it’s luxury, it’s a digital rights imperative. That is really, really, really true. So, while awareness on universal acceptance is important, Universal Acceptance has grown, as you have said. Adoption remains uneven, for sure. So, the gaps are very large. For instance, only 26% of tested email servers today support email addresses in internationalized scripts. Imagine that, 26% only. There was a study in 2025 by the Universal Acceptance Steering Group. It found that the majority of local websites still reject email addresses in form fields. Overall, I’m very sorry to say, but universal acceptance readiness is the exception. It’s an exception to find that in systems. And that is sad. I find it very sad to see my language is not recognized in the digital world in 2025. I feel my culture is not recognized. It’s just sad, in a way. What we face is not lack of standards. I want to highlight that. It’s not that we lack the standards. The standards are there. What we lack is consistent implementation across code, procurement, and policy. Now, how can we raise awareness, to go to your question, and accelerate action? To close the universal acceptance readiness gap, we need to target three key communities. The developers and technology providers, the service providers, and the policymakers. How are we raising awareness? One of our key awareness activities that we launched to close this gap is the Universal Acceptance Day events. They started in 2023, and it has grown into a global awareness campaign. UNESCO is an invaluable partner in this awareness campaign, and many others, but I will not go into details. As I mentioned, we need three key categories. What do they need to do? Developers, they need to integrate universal acceptance and email address internationalization into coding, best practices. They have to make it part of a core curricula in software engineering, and they can use testing tools, which are already available from the universal acceptance student group. The service providers… They need to conduct universal acceptance readiness audits of platforms and websites and have to fix form validation errors. Policymakers, very important to my mind. They can include universal acceptance in public procurement guidelines and nationally government service design. A government should lead by example. If a citizen cannot register with their local script email, we have failed at inclusivity. Governments can generate demand. They have this power through procurement. And by being demand creators, they can pressure for providers to upgrade. This is something we need. Providers do not feel the need to upgrade. When it comes to the second part of your question, Davide, which is about technical and policy priorities in the next phase, on the technical front, obviously, I sort of covered that we need to increase support for email address internationalization, readiness, and quotas platforms to test and fix errors. On the regulatory front, we have to incentivize universal acceptance through standard bodies, international digital cooperation instruments like the Global Digital Compact, which was recently adopted. It does reference that we need to push through. Create cross-sector incentives. As I said, for example, digital public infrastructure funding tied to universal acceptance readiness. We also have to support the partnerships between registries and indigenous communities and developers to create the main domains and email services in underrepresented languages. Dr. Uwaweru was talking about the more than 2,000 local languages that we have in Africa. This is what I mean. How do they have an online presence? Except the content that is in a language or not, but the email address, the website. On that point, I will say next year, there will be the launch of the new… and Mr. David Storti. I would like to start by saying that this is the first of two rounds of domain names in 2026 by ICANN, and this will give an opportunity to enable more domains in more scripts. People can come and ask for domain names in their own local script. As Guilherme said at the beginning, it involves many different angles. It takes a collaborative way forward. We’re very proud of our collaboration with UNESCO on that front, and I think that is bearing fruit. Let me end with this. Multilingualism online must move from being an aspiration to a system default. Let’s make the universal acceptance the norm, not the exception. Let’s normalize inclusion, not just promote it. Thank you.

Davide Storti: Thank you very much for this intervention. I’d like to thank all the speakers, because there is a lot to say on this topic, and particularly the last one we heard. I mean, just think about, you know, the hard work that has been put into this. I mean, it’s not easy. It’s not easy. It’s not easy. I mean, you can think about, you know, how to upgrade the millions and millions of forms, apps and interfaces which are already there. It’s a daunting task, so it’s not something which happens overnight. Unfortunately, I’m really sorry. Apologies from my side. There is not really much time for questions. So I, with your indulgence, I would like to give the floor to Mr. Dr. Tawfik Jelass for some final remarks.

Tawfik Jelassi: Thank you, Davide. I will use only ten seconds, not more than that, to thank all the panelists for their insights and perspectives for this very important topic. Clearly, this is top of the agenda going forward, whether it is WSIS plus 20 plus plus or whether the IGF for the next decade or whether the Global Digital Compact, I think multilingualism in cyberspace, promoting linguistic diversity should be a priority for all stakeholders in an inclusive manner. Final observation. I couldn’t resist that we had 30 people in the audience, 20 women and less than 10 men. What does that say, is multilingasm more important for women then men? Is this how we can do an inclusive approach, to promote linguistic diversity? It did strike me when I saw that in the room. Hopefully, men can also play their part, its not a matter of gender it is a matter of we need all, to bring all, less herd voices in cyberspace.

Davide Storti: Thank you very much. Thank you very much and just to conclude I would like to thank my colleagues who are not with us today but are online, David Castillo Parra and Flora Udenzik for their support in organizing this session. Thank you very much and I wish Action Line C8 a great future and if you have suggestions for sharing ahead of the 20 years review, I think it’s very important that you share with me, come please, because we have a message to send to the review and that’s your voice is important, our voice is important. So thank you very much.

Agreement points

Digital language divide represents a fundamental barrier to inclusive digital societies

Speakers

– Tawfik Jelassi
– David Waweru
– Elena Plexida
– Valts Ernstreits

Arguments

Only 120 out of 8,000 world languages are online, with 91% of content in just 14 languages

African languages represent less than 0.1% of websites despite Africa having 2,000+ languages

Only 26% of email servers support internationalized email addresses, making digital identity invalid for many users

Limited data availability prevents AI solutions for endangered languages like Livonian with fewer than 20 speakers

Summary

All speakers acknowledge that the current digital infrastructure severely underrepresents linguistic diversity, creating systemic barriers for billions of users who cannot access digital services in their native languages

Topics

Multilingualism | Digital access | Cultural diversity

Community-led approaches are essential for successful language technology development

Speakers

– Valts Ernstreits
– Sofiya Zahova
– David Waweru

Arguments

Speaker involvement programs hiring native speakers for digital resource development creates both technological advancement and language revitalization

Community-led initiatives are most impactful when culturally grounded and supported by long-term partnerships

African grassroots organizations and local startups are creating voice recognition and translation for overlooked languages

Summary

Speakers agree that the most effective language preservation and technology development occurs when communities themselves lead the initiatives, contributing their linguistic expertise and cultural knowledge

Topics

Multilingualism | Cultural diversity | Capacity development

Multi-stakeholder collaboration is necessary to address linguistic diversity challenges

Speakers

– Tawfik Jelassi
– Guilherme Canela de Souza Godoi
– Elena Plexida

Arguments

Multi-stakeholder collaboration involving governments, communities, private sector, and civil society is essential for progress

The UN International Decade of Indigenous Languages (2022-2032) has 15 countries with national action plans and 20 more developing them

Governments should lead by example through procurement guidelines and public service design that supports all scripts

Summary

Speakers emphasize that no single actor can solve the digital language divide alone, requiring coordinated efforts across governments, communities, private sector, and international organizations

Topics

Multilingualism | Legal and regulatory | Cultural diversity

Similar viewpoints

Both speakers emphasize that linguistic diversity in digital spaces requires systematic institutional support and policy implementation rather than just technical solutions

Speakers

– Sofiya Zahova
– Elena Plexida

Arguments

Multilingual accessibility should be treated as critical infrastructure like healthcare, requiring state support

Universal acceptance requires consistent implementation across code, procurement, and policy rather than new standards

Topics

Multilingualism | Digital standards | Legal and regulatory

Both speakers highlight the importance of policy frameworks and international cooperation while acknowledging implementation gaps in existing initiatives

Speakers

– David Waweru
– Guilherme Canela de Souza Godoi

Arguments

African Union’s Smart Africa Coalition launched AI for Africa Blueprint involving 40 nations

UNESCO’s 2003 recommendation on multilingualism requires government reporting, but only 60% of reporting countries implement policies

Topics

Multilingualism | Legal and regulatory | Digital standards

Both speakers provide concrete examples of successful community-driven digital initiatives that preserve and promote linguistic and cultural diversity

Speakers

– Sofiya Zahova
– David Waweru

Arguments

Indigenous communities like Sámi and Inuit have successfully developed language technologies under their own leadership

Digital platforms like African Storybook provide culturally relevant content in 40+ African languages

Topics

Multilingualism | Cultural diversity | Online education

Unexpected consensus

Technical implementation challenges are more significant than lack of standards

Speakers

– Elena Plexida
– Valts Ernstreits

Arguments

Universal acceptance requires consistent implementation across code, procurement, and policy rather than new standards

Limited data availability prevents AI solutions for endangered languages like Livonian with fewer than 20 speakers

Explanation

Both speakers, coming from very different contexts (global internet governance vs. endangered language preservation), agree that the main barriers are implementation and resource challenges rather than lack of technical standards or solutions

Topics

Digital standards | Multilingualism | Digital access

Government procurement can drive market change for linguistic inclusion

Speakers

– Elena Plexida
– Guilherme Canela de Souza Godoi

Arguments

Governments should lead by example through procurement guidelines and public service design that supports all scripts

UNESCO’s 2003 recommendation on multilingualism requires government reporting, but only 60% of reporting countries implement policies

Explanation

The consensus on government procurement as a lever for change is unexpected given that one speaker focuses on technical standards while the other on policy frameworks, yet both see government demand as crucial for market transformation

Topics

Legal and regulatory | Digital standards | Multilingualism

Overall assessment

Summary

Strong consensus exists on the severity of the digital language divide, the necessity of community-led approaches, and the need for multi-stakeholder collaboration. Speakers agree on both the problem diagnosis and general solution approaches.

Consensus level

High level of consensus with complementary perspectives rather than conflicting viewpoints. This strong agreement suggests a mature understanding of the challenges and potential for coordinated action across different sectors and regions.

Different viewpoints

Approach to technology development for endangered languages

Speakers

– Valts Ernstreits
– Elena Plexida

Arguments

Limited data availability prevents AI solutions for endangered languages like Livonian with fewer than 20 speakers

Universal acceptance requires consistent implementation across code, procurement, and policy rather than new standards

Summary

Valts emphasizes the fundamental challenge of data scarcity preventing AI development for endangered languages, while Elena focuses on implementing existing standards rather than developing new technologies. They represent different perspectives on whether the priority should be creating new language-specific solutions or implementing universal technical standards.

Topics

Multilingualism | Digital standards | Cultural diversity

Role of major technology companies in language inclusion

Speakers

– David Waweru
– Sofiya Zahova

Arguments

African grassroots organizations and local startups are creating voice recognition and translation for overlooked languages

Indigenous communities like Sámi and Inuit have successfully developed language technologies under their own leadership

Summary

While both advocate for community-driven approaches, David emphasizes partnerships between big tech and grassroots organizations as beneficial, while Sofiya warns against extractive digitization and emphasizes that communities must be authors of their digital futures, showing more skepticism toward external involvement.

Topics

Multilingualism | Cultural diversity | Digital access

Unexpected differences

Concerns about major platform inclusion efforts

Speakers

– Sofiya Zahova

Arguments

Roma communities show remarkable digital engagement despite systematic barriers, with initiatives digitizing 19th-century Romani literary heritage

Explanation

Unexpectedly, Sofiya raised concerns about Google Translate’s inclusion of Lach Romani, questioning the methodology and lack of community consultation. This represents an unexpected disagreement with what might typically be seen as progress, highlighting that inclusion efforts can be problematic when done without proper community involvement.

Topics

Multilingualism | Cultural diversity | Human rights principles

Overall assessment

Summary

The discussion showed remarkably high consensus on goals with subtle but important disagreements on implementation approaches. Main areas of disagreement centered on the balance between top-down versus bottom-up approaches, the role of major technology companies, and whether to focus on new technology development versus implementing existing standards.

Disagreement level

Low to moderate disagreement level with high strategic significance. While speakers agreed on fundamental goals of linguistic inclusion and cultural preservation, their different approaches could lead to competing resource allocation and policy priorities. The disagreements reflect deeper tensions between community autonomy and institutional efficiency in digital language preservation efforts.

Partial agreements

Similar viewpoints

Both speakers emphasize that linguistic diversity in digital spaces requires systematic institutional support and policy implementation rather than just technical solutions

Speakers

– Sofiya Zahova
– Elena Plexida

Arguments

Multilingual accessibility should be treated as critical infrastructure like healthcare, requiring state support

Universal acceptance requires consistent implementation across code, procurement, and policy rather than new standards

Topics

Multilingualism | Digital standards | Legal and regulatory

Both speakers highlight the importance of policy frameworks and international cooperation while acknowledging implementation gaps in existing initiatives

Speakers

– David Waweru
– Guilherme Canela de Souza Godoi

Arguments

African Union’s Smart Africa Coalition launched AI for Africa Blueprint involving 40 nations

UNESCO’s 2003 recommendation on multilingualism requires government reporting, but only 60% of reporting countries implement policies

Topics

Multilingualism | Legal and regulatory | Digital standards

Both speakers provide concrete examples of successful community-driven digital initiatives that preserve and promote linguistic and cultural diversity

Speakers

– Sofiya Zahova
– David Waweru

Arguments

Indigenous communities like Sámi and Inuit have successfully developed language technologies under their own leadership

Digital platforms like African Storybook provide culturally relevant content in 40+ African languages

Topics

Multilingualism | Cultural diversity | Online education

Key takeaways

There is a critical digital language divide with only 120 out of 8,000 world languages represented online, and 91% of content available in just 14 languages

Multilingualism in digital spaces requires systemic change involving policy frameworks, community-driven approaches, technical implementation, and multi-stakeholder collaboration

Community involvement is essential for successful language technology development, with native speakers playing crucial roles in creating culturally appropriate digital resources

Universal acceptance of internationalized domain names and email addresses is a digital rights imperative, but currently only 26% of email servers support non-Latin scripts

Cultural and creative industries must be integrated with linguistic diversity efforts to preserve indigenous knowledge systems and storytelling traditions

Government leadership through procurement policies and public service design can drive demand for multilingual digital infrastructure

Resource constraints and limited data availability pose significant challenges for endangered languages in developing AI solutions and digital presence

Resolutions and action items

UNESCO will continue the consultation process for the global roadmap on language technologies and multilingualism until July 12th

Governments should be encouraged to report back on their implementation of the 2003 UNESCO recommendation on multilingualism in cyberspace

The UN International Decade of Indigenous Languages should be leveraged as a window of opportunity, with 20 additional countries developing national action plans

ICANN will launch new internationalized domain names in 2026, providing opportunities for more domains in local scripts

Universal Acceptance Day events should continue as a global awareness campaign targeting developers, service providers, and policymakers

Multi-stakeholder collaboration should be strengthened involving governments, communities, private sector, civil society, and technical communities

Participants should provide input for the WSIS+20 review process to ensure multilingualism remains a priority

Unresolved issues

How to effectively scale community-driven approaches to the thousands of underrepresented languages globally

Addressing the resource and capacity constraints that prevent many endangered language communities from entering the digital domain

Ensuring ethical considerations and avoiding extractive digitization when developing language technologies for indigenous communities

Bridging the gap between technical standards for universal acceptance and consistent implementation across millions of existing applications and websites

Balancing the need for AI language models with limited data availability for low-resource languages

Determining appropriate representation and consultation processes when selecting language variants for digital platforms

Measuring and tracking progress in multilingual digital inclusion ahead of the WSIS+20 review

Suggested compromises

Treating multilingual accessibility as critical infrastructure requiring state support similar to healthcare, while recognizing resource limitations

Combining top-down policy frameworks with bottom-up community-driven initiatives to address different aspects of the language divide

Leveraging partnerships between global technology companies and grassroots organizations to balance technical capabilities with cultural sensitivity

Using government procurement power to create market demand for universal acceptance while providing technical support for implementation

Focusing on both language preservation and unlocking cultural heritage through technology, recognizing that communities have different priorities and needs

Think of the digital world as a vast library. If only a few languages are represented by books on the shelves of that library, think how many civilizations, how many histories would be left unheard and unread. We must ensure that this library reflects the full spectrum of human experience, where every language has a place, every voice can be heard, and every culture can contribute.

Speaker

Tawfik Jelassi

Reason

This metaphor powerfully reframes the technical challenge of multilingualism as a fundamental issue of human representation and cultural preservation. It transforms the discussion from technical implementation to human rights and cultural equity.

Impact

This metaphor set the philosophical foundation for the entire discussion, with subsequent speakers consistently returning to themes of inclusion, representation, and cultural preservation rather than purely technical solutions.

So basically if you do not able to produce any digital data so you cannot have any digital technology at all and this is true for quite many world’s languages who cannot simply enter digital domain or the same problem that we have in Livonian case if we have limited data it’s very hard to create AI solutions or large language models.

Speaker

Valts Ernstreits

Reason

This comment reveals a fundamental paradox in digital inclusion – languages need digital presence to benefit from technology, but they need resources to achieve digital presence. It highlights how current AI development inadvertently excludes endangered languages.

Impact

This shifted the conversation from discussing solutions to acknowledging structural barriers, leading other speakers to address community-driven approaches and the need for alternative methodologies for low-resource languages.

We must avoid the trap of extractive digitization and ensure that communities, especially indigenous and minority ones, are the authors of their digital futures, and they decide what has to be made available online and how digital technologies have to be utilized.

Speaker

Sofiya Zahova

Reason

This comment introduces the critical concept of ‘extractive digitization’ and emphasizes community agency over external technological imposition. It challenges the assumption that any digitization is inherently beneficial.

Impact

This comment elevated the discussion to address power dynamics and self-determination, influencing subsequent speakers to emphasize community leadership and ethical considerations in their examples and recommendations.

Universal acceptance is not the technical luxury, it’s a digital rights imperative… In human terms, what does that mean? A person’s digital identity, tied to a language or a script, is rendered invisible or invalid in that way.

Speaker

Elena Plexida

Reason

This reframes a technical standard as a human rights issue, making the abstract concept of universal acceptance tangible by connecting it to personal digital identity and belonging.

Impact

This comment bridged the technical and human aspects of the discussion, demonstrating how seemingly minor technical details have profound implications for digital inclusion and cultural recognition.

More websites use Norwegian, spoken by about 4 million people, than use Swahili, which is spoken by over 200 million people. A stark illustration of the digital linguistic divide.

Speaker

David Waweru

Reason

This striking statistical comparison exposes the disconnect between linguistic demographics and digital representation, revealing how digital spaces don’t reflect real-world language use patterns.

Impact

This concrete example provided powerful evidence for the systemic nature of digital linguistic inequality, reinforcing earlier theoretical discussions with hard data that other speakers could reference.

I couldn’t resist that we had 30 people in the audience, 20 women and less than 10 men. What does that say, is multilingualism more important for women than men?

Speaker

Tawfik Jelassi

Reason

This unexpected observation about gender representation in the audience introduces an intersectional dimension to multilingualism, suggesting that different groups may have varying stakes in linguistic diversity.

Impact

Though made at the end, this comment opened up new questions about who is most affected by linguistic exclusion and who advocates for multilingual inclusion, adding complexity to the understanding of stakeholder engagement.

Overall assessment

These key comments transformed what could have been a technical discussion about language technology into a profound exploration of digital rights, cultural preservation, and systemic inequality. The metaphor of the digital library established an inclusive framework that subsequent speakers built upon, while the technical insights about data requirements and universal acceptance grounded the discussion in practical realities. The emphasis on community agency and the stark statistical comparisons provided both ethical guidance and empirical evidence for the urgency of the issue. Together, these comments created a multi-layered conversation that addressed technical, cultural, ethical, and political dimensions of multilingualism in digital spaces, ultimately framing it as a fundamental aspect of human dignity and cultural survival rather than merely a technical challenge to be solved.

How can we improve government reporting on multilingualism policies under the 2003 UNESCO recommendation?

Speaker

Guilherme Canela de Souza Godoi

Explanation

Only 60% of governments that report back are actually implementing policies in this area, and many governments don’t report at all. This is critical for understanding what’s happening and improving policies.

How can we engage more effectively with the International Decade of Indigenous Languages survey and initiatives?

Speaker

Guilherme Canela de Souza Godoi

Explanation

There’s a current survey open until the weekend about indigenous languages worldwide, and this represents a window of opportunity that ends in 2032 to move the needle on multilingualism technology.

How can we develop ethical frameworks and address bias in AI language technology models?

Speaker

Sofiya Zahova

Explanation

Iceland is collaborating with UNESCO on a global initiative focusing on ethical aspects and bias questions in language technology and AI, which is crucial for fair representation.

How can we ensure transparent consultation and community participation in language technology decisions?

Speaker

Sofiya Zahova

Explanation

The inclusion of Lach Romani in Google Translate raised concerns about representation and methodology for selecting one dialect over others without proper community consultation.

How can we create a global framework for inclusion of low-resource languages in AI design and evaluation?

Speaker

Sofiya Zahova

Explanation

There’s a pressing need for international cooperation to ensure underrepresented cultures are included in digital technology development, particularly AI systems.

How can we implement the African Union’s Artificial Intelligence for Africa Blueprint effectively?

Speaker

David Waweru

Explanation

While strategic roadmaps exist, the key challenge is implementing these major roadmaps across 40 African nations to address digital language divides.

How can we increase universal acceptance readiness from the current 26% of email servers?

Speaker

Elena Plexida

Explanation

Only 26% of tested email servers support email addresses in internationalized scripts, and universal acceptance readiness remains an exception rather than the norm.

How can governments integrate universal acceptance into public procurement guidelines?

Speaker

Elena Plexida

Explanation

Governments can lead by example and create demand through procurement, pressuring providers to upgrade their systems to support multilingual digital identities.

How can we upgrade millions of existing forms, apps and interfaces to support universal acceptance?

Speaker

Davide Storti

Explanation

The technical challenge of retrofitting existing digital infrastructure to support multilingual domain names and email addresses is daunting and requires systematic approach.

Why is there a gender disparity in multilingualism advocacy and how can we engage more men in this cause?

Speaker

Tawfik Jelassi

Explanation

The observation that 20 women and less than 10 men attended the session raises questions about whether multilingualism is perceived differently by gender and how to achieve more inclusive participation.