Risks of Over-Relying on AI-Powered Content Moderation Systems in Diverse Cultural Contexts

Key Risks Identified

Cultural and Linguistic Bias

Multiple sessions highlighted the fundamental issue of cultural bias in AI-powered content moderation systems. As noted in the Parliamentary Session on Digital Deceit, “At a fundamental level, the AI model that you see today are trained on English speaking data from the internet that is mainly generated in the global north or in the north. And what it means is that it really sort of aggregate a specific vision of the world with its inherent biases.”

The workshop on Generative AI in Content Moderation provided concrete examples of these biases, with speakers noting that “removing terrorist content in Arabic got it wrongly 77% of the time” and how “Instagram falsely flagged Al-Aqsa Mosque, which is the third holiest mosque in Islam as a terrorist organization.”

Language Diversity Challenges

Several sessions emphasized how AI systems struggle with linguistic diversity. In the workshop on AI Readiness in Africa, speakers highlighted that “Remember, we have more than 2,000 languages in Africa. So if we allow those languages to be trained on the AI system, but not trained by us, it basically means even indigenous people will be impacted by any cultural biases using AI.”

The Parliamentary Session on Protecting Vulnerable Groups further illustrated this challenge, noting that “in our context, for example, in Uganda, there are about 50 languages that are spoken, in Uganda alone, not considering the whole continent. And because these are smaller countries, they don’t have a huge market share, you know, on these online platforms. They’re often not prioritized.”

Cultural Sensitivity and Context Understanding

The Parliamentary Session on Digital Policy Practices highlighted critical cultural sensitivity issues, with speakers explaining that “So we need to be very careful of the fact that the culture in which we are living, the social media platforms have to be sensitive about that culture. And this is the real challenge, how to make the social media platforms sensitive to the cultures.”

Systematic Discrimination and Over-Moderation

The workshop on International Law in Digital Spaces documented systematic issues, noting “Systematic censorship and discriminatory content moderation policies by these platforms, as seen in the suppression of Palestinian voices, undermine these digital rights. And the disproportionate over-moderation leads to restrictions limiting the reach of Palestinian content at the international level.”

Proposed Solutions and Approaches

Regional Expertise and Local Context

The High-Level Session on Information Space highlighted the importance of regional expertise, with TikTok’s representative explaining: “Sitting in North America, I don’t know the right solutions for Estonia. I can’t see what’s coming around the corner. And so it’s critically important that we have these regional safety councils with people who do have that expertise and will bring it to us and share it with us so that we can try to solve problems before they find their ways onto the platform.”

Localized AI Development

The session on Ethical AI-Generated Content provided a concrete example of localized solutions, with speakers describing how they are “working on our own AI models that are designed to classify hate speech and violence on social media platforms and different chat platforms in two different languages, Hebrew and Arabic” with “words, terms, definitions and data labeling that reflect the specific region we live in and countries we work on.”

Community Participation and Human Oversight

The workshop on AI in Content Moderation emphasized community involvement, suggesting to “engage with greater community leadership and participation in the building of LLM” and ensuring that human feedback processes involve affected communities rather than just Silicon Valley experts.

Hybrid Approaches

Several sessions advocated for combining AI with human oversight. The High-Level Session noted that “When we pair technology with human moderation, we have found that we are able to remove violative content 98% of the time, proactively, before it’s reported to us as a platform.”

Multistakeholder Approach

The Parliamentary Session on Digital Deceit emphasized the need for “multi-stakeholder perspective, but also from different local and regional level to input diversity in every stage of the AI life cycle from the data to the outputs.”

Conclusion

The discussions across multiple IGF 2025 sessions revealed that over-reliance on AI-powered content moderation systems poses significant risks in diverse cultural contexts, including systematic bias, linguistic discrimination, and cultural insensitivity. The proposed solutions emphasize the need for localized approaches, community participation, human oversight, and multistakeholder engagement to address these challenges effectively.

Long-term Implications of Private Digital Platforms in Shaping Public Discourse and Democratic Processes

The Internet Governance Forum 2025 discussions revealed deep concerns about the growing power of private digital platforms and their impact on democratic governance. Multiple sessions highlighted how a small number of tech companies now control global information flow and democratic discourse.

Concentration of Power

A central theme across sessions was the unprecedented concentration of power in the hands of a few tech giants. In the Dynamic Coalition Collaborative Session, audience member Kjetil Kjernsmo stated that “the power of internet governance is not in this room. It is chiefly with big tech.” Dr. Rajendra Pratap Gupta reinforced this concern, noting that “it is small number of large companies that drive the internet rather than large number of small companies.”

The Parliamentary Session 4 emphasized this power imbalance, with Anna Luhrmann from Germany stating that “the actual power in the internet currently resides with the big tech companies that have in their respective fields quasi-monopolies in many areas.”

Threats to Democratic Processes

Multiple sessions documented how platforms pose existential threats to democratic institutions. In the Platform Governance and Duty of Care Workshop, Beatriz Kira identified critical threats including “disinformation campaigns that consume elections, hate speech that spills from screens into the streets and real-world harms, and co-ordinated harassment that could silence marginalised voices.”

The Closing Ceremony featured Maria Ressa’s stark warning about platforms’ transformation from democratic tools to weapons of oppression. She observed that “In my own country, in the Philippines, I’ve watched social media transform from a tool of liberation into a weapon of oppression.” She emphasized that “Algorithms that amplify our worst impulses, that reward outrage over empathy, that trap us in bubbles of our own biases, these are not inevitable. They’re choices.”

Business Model Concerns

Several sessions identified platforms’ advertising-driven business models as fundamentally problematic for democratic discourse. In the Climate Change Information Integrity Open Forum, Harriet Kingaby explained how “advertising is essentially the funding model behind the attention economy” creating “unhealthy incentives for the production of content that has devastating consequences for information integrity.”

The Parliamentary Session on Digital Deceit reinforced this concern, with Camille Grenier noting that “big tech business models prioritize their… have laid out monetization for profit. And these business models create dependencies for private and public organizations, as well as individuals, and facilitate the weaponization of information.”

Impact on Media and Journalism

The discussions revealed how platforms undermine traditional journalism and media independence. In the High Level Session on Losing the Information Space, Minister Jaffery explained that “news media’s ability to perform their function as watchdogs and providers of reliable information are challenged by big tech platforms.”

The Information Integrity Workshop highlighted the scale of platform dominance, with Beatriz Barbosa noting that “two companies, Google and Meta, for sure, hold a dominant global position in the distribution of news and information.”

Algorithmic Manipulation and Content Moderation

Sessions examined how algorithmic systems shape public discourse in problematic ways. The Truth Under Siege event featured Pavel Zoneff explaining that “it’s really the issue of a handful of big tech companies that now mediate the flow of nearly all information and their algorithms. Their algorithms and platforms really rule and decide what is visible at scales.”

The AI Content Moderation Workshop warned about increasing homogenization, with Marlene Owizniak explaining that content moderation decisions “defined at the foundation level will also be replicated on the deployment one and really there’s even more homogeneity of speech as before.”

Impact on Youth and Democratic Participation

Several sessions highlighted particular concerns about platforms’ influence on young people’s political engagement. The Youth in Public Discourse Lightning Talk raised critical questions about democratic participation, with Katarina Juni Moneta asking: “So what happens to democracy when young people spend more time listening to influencers than to journalists and politicians?”

The High Level Session 5 documented specific harms, with Minister Tung noting concerns about platform algorithms where “Our kids are screaming for help because they are having trouble with sleep, with health issues, body issues, and so forth, because of these algorithms.”

Global South Perspectives

Discussions revealed how platform dominance particularly affects Global South countries. The Connectivity Gap Workshop highlighted “the regulatory power imbalances that now are growing between states and big tech companies. Countries of the global majority have today a minimum capacity to demand the fulfillment of guarantees and rights of these companies.”

Calls for Democratic Reclamation

Many sessions called for reclaiming democratic control over digital spaces. In the High Level Session 1, Thibaut Bruttin argued that <a href="https://dig.watch/event/internet-governance-forum-2025/high-level-session-1-losing-the-information-space-ensuring-human-rights-and-resilient-societies-in-the-age-of-big-tech-2?diplo-deep-link-text=It%27s%20a%20fantasy%20to%

Based on the available session transcripts from the Internet Governance Forum 2025, the specific question regarding frameworks for ensuring content moderator well-being was not directly addressed or discussed in any of the sessions. While various sessions covered related topics such as digital governance, AI regulation, online safety, cybersecurity, and digital cooperation, none specifically focused on the mental health challenges, ethical dilemmas, and support systems needed for content moderators who work to maintain healthier digital environments.

The sessions that came closest to addressing content moderation issues included discussions on combating sexual deepfakes, protecting vulnerable groups online, and building safe online environments, but these did not delve into the human cost and welfare concerns of those responsible for moderating harmful content.

This represents a significant gap in the IGF 2025 discussions, as content moderator well-being is a critical aspect of maintaining digital safety and requires urgent attention from policymakers, tech companies, and international organizations to develop comprehensive support frameworks.

What is missing in our approaches to addressing the environmental impact of digital technologies?

The environmental impact of digital technologies emerged as a significant concern across multiple sessions at IGF 2025, with discussions revealing several critical gaps in current approaches to addressing these challenges.

Lack of Integration in Digital Governance Frameworks

A fundamental issue identified was the absence of environmental considerations in digital policy frameworks. As noted in the ROAM X session, Fabio Senne observed that “environmental issues are largely overlooked in digital policies so far” and that “there are still issues such as energy consumption, e-waste and emissions that are not yet well integrated into the governance framework.”

Massive AI Energy Consumption and Water Usage

Multiple sessions highlighted the alarming energy and resource consumption of AI systems. In the AI for Sustainable Development session, Professor Rony Medaglia revealed that “to generate one single image with a large language model, such as chat GPT, uses the same amount of CO2 as charging your mobile phone up to 50%” and that “the global AI demand may be accountable in two years from now for a water withdrawal equal to six times the whole country of Denmark for a whole year of water.”

Lack of Transparency and Measurement

A critical gap identified was the absence of transparent reporting on environmental impacts. In the Make Your AI Greener workshop, Ioanna Ntinou emphasized that “we should start by having some transparency in the way we report energy” and called for “developing evaluation standards at prompt level.”

This transparency deficit was further highlighted in the AI at a Crossroads session, where Ana Valdivia noted that “84% of widely used large language models provide, no disclosure at all of their energy use or emissions.”

E-Waste Management Crisis

The E-Waste Management session revealed significant gaps in waste handling, with Emmanuel Niyikora noting that “we have 32%, which is formally collected and recycled. So which means the remaining 78 remain unmanaged.” A participant emphasized the measurement challenge: “We can’t manage what we can’t measure. And right now, the global e-waste system is mostly blind.”

Climate Justice and Global Inequality

Environmental impacts disproportionately affect the Global South. In Main Session 2, Anna from R3D Mexico highlighted how companies move data centers to avoid environmental scrutiny: “when it has established hyperscale data centers in places like the Netherlands that had made people publicly pressured for them to stop being constructed there, so then you move them to global south countries” causing “all the issues with extractivism, with hydric crisis, with pollution arrive to other communities.”

Need for Knowledge-Based Governance

The Sustainable Digital Growth session identified fundamental research gaps. Jan Eivind Velure emphasized that “the governance must be knowledge-based” and called for comprehensive lifecycle analyses to establish “a shared reference point that we can start talking around.”

Misaligned Incentives and Investment Priorities

The Local AI Policy Pathways session highlighted problematic investment patterns, with Anita Gurumurthy noting that “Between 2022 and 2025, AI-related investment doubled from $100 to $200 billion. By comparison, this is about three times the global spending on climate change adaptation.”

In the Make Your AI Greener workshop, Mark Gachara observed that “there’s a lot of money going into climate or smart agriculture or climate AI solutions But it’s to make them more efficient more effective as opposed to how do we mitigate? The harms that they are causing.”

Children’s Environmental Concerns

Young people demonstrated heightened environmental awareness. In the Elevating Children’s Voices session, Dr. Mhairi Aitken reported that “Where children have awareness or access to information about the environmental impacts of generative value models, they often choose not to use those models.”

Need for Active Intervention

Multiple sessions emphasized that environmental sustainability will not emerge naturally. In the Make Your AI Greener workshop, Ioanna Ntinou stressed that “sustainable AI is not going to emerge by default. We need to incentivize and it needs to be actively supported.”

The discussions revealed that addressing the environmental impact of digital technologies requires comprehensive policy integration, transparent reporting mechanisms, better measurement systems, climate justice considerations, knowledge-based governance frameworks, aligned incentives, and active intervention rather than relying on market forces alone.

No relevant discussions found.

Designing and Enforcing Gender-Responsive Laws and Legal Frameworks for Women’s Online Protection

The discussions across multiple IGF 2025 sessions revealed significant challenges and emerging solutions for protecting women from online harm while promoting their digital participation. The conversations highlighted both the urgent need for comprehensive legal frameworks and the complexity of implementing effective protections.

The Scale of the Challenge

Sessions consistently emphasized the growing nature of online gender-based violence. In Parliamentary Session 3, Nighat Dad shared alarming statistics from Pakistan: “Since 2016, through our digital security helpline, we have dealt with more than 20,000 complaints from hundreds of young women every month, female journalists, now more from women influencers and content creators, women politicians, scholars, and students.” The digital divide compounds these challenges, as noted in High Level Session 5 by Baroness Jones: “Globally there were 244 million more men than women using the internet in 2023.”

Cultural Sensitivity in Legal Design

Multiple sessions emphasized the critical importance of cultural context in designing gender-responsive laws. In Parliamentary Session 5, Anusha Rahman Ahmad Khan from Pakistan provided a stark example: “And in my country, even an aspersion on a girl is good enough to kill her. So even if they would not die physically, but they are dead emotionally. So we need to be very careful of the fact that the culture in which we are living, the social media platforms have to be sensitive about that culture.” This sentiment was echoed in Open Forum #17, where Amira Saber highlighted deepfake risks: “Imagine a girl who is living in a village with certain cultural norms and she has leaked photographs of her on pornography. This might threaten her life directly. She could be killed.”

Challenges with Existing Legal Frameworks

Sessions revealed significant problems with current approaches to gender-responsive legislation. In Parliamentary Session 3, Neema Iyer highlighted a critical paradox: “the laws that do exist, especially in our context, have actually been weaponized against women and marginalized groups. So many of these, you know, cybercrime laws or data protection laws, have been used against women, have been used against dissenting voices, against activists, to actually punish them rather than protect them.” She further noted that “legislative frameworks are often too narrow. They, you know, they focus on takedowns or criminalization, or they borrow from Western contexts, but they don’t really meet the lived realities of women.”

Enforcement Challenges

Even where appropriate laws exist, enforcement remains problematic. In Networking Session #232, speaker Oni Kamakwakwa identified a critical gap: “we always have really great policies, cybersecurity laws that criminalize the behavior. However, there is no knowledge of how to enforce it, including within the judiciary as well… we have women who continuously get harassed, including serious cases of revenge porn, where even the police are laughing at the incidents because they don’t realize that, in the same way you see someone speeding down the highway and you chase them because they’ve broken a law, this is how we expect judiciary clusters to actually respond to these cases.”

Judicial challenges were further explored in WS #190, where Marin Ashraf noted: “Online gender-based violence is typically dealt as a criminal offense in India under the General Penal Code and the Information Technology Act… In our study, we found that in several cases of online gender-based violence, it unfortunately failed to meet the high burden of proof that is required because of difficulties in bringing in expert testimony, relying on witnesses, and ensuring the admissibility of digital evidence.”

Emerging Technologies and New Challenges

The rise of AI and deepfake technology presents new challenges for legal frameworks. In WS #70, Yi Teng Au explained the legal complexity: “in many countries, the laws against deepfakes is very muddy because it’s a picture that is modified. Although it’s a picture of a face, the body, it’s not. So a lot of the times, it’s very hard to persecute the perpetrators.” Juliana Cunha emphasized the gendered nature of this issue: “This is a problem about gender-based violence and it’s just another form of it” reflecting “broader systemic gender inequalities.”

Promising Legislative Developments

Several sessions highlighted positive legislative developments. In Parliamentary Session 3, Raoul Danniel Abellar Manuel described Philippine efforts: “we recently had the Republic Act 11930… aside from content taken one major component of this is the assertion of extraterritorial jurisdiction” and “the House of Representatives, on its part, approved the expanded anti-violence bill. It defines psychological violence, including different forms, including electronic or ICP devices.”

The UK’s approach was mentioned in WS #70, where Kenneth Leung noted the <a href="https://dig.watch/event/internet-governance-forum-2025/ws-70-comb

Potential Negative Consequences of Individual-Centric Digital Rights Framing

Several sessions at the Internet Governance Forum 2025 addressed the limitations and potential negative consequences of framing digital rights primarily through an individual liberties lens, highlighting the need to consider collective rights, public interest, and broader social responsibilities.

The Dominance of Individual Rights Over Collective Interests

The discussion revealed a clear pattern where individual rights have been prioritized at the expense of collective and public interest considerations. At the New Technologies and the Impact on Human Rights session, Allison Gilwald noted that “the individual rights have taken preference over maybe collective rights or public interest rights” and emphasized “extending those rights to look at some of the kind of collective implications of that.”

Invisibility of Non-Users and Affected Communities

One significant consequence of individual-focused rights frameworks is the exclusion of those who are indirectly affected by digital technologies but are not direct users. Anita Gurumurthy emphasized that “we need to account for those who are non-users. People who mine rare earth minerals, those whose lands are given away to data centers, who are dispossessed.”

Need for Collective Data Rights

The limitations of individual data rights were particularly highlighted in discussions about data governance. At the Local AI Policy Pathways session, Sarah Nicole argued that “The question of having a stake in your data has often been framed on a personal level… So the answer will not be on an individual perspective, but it would be on a collective one.” She emphasized that “the mentality really needs to shift from this personal data frame discussion… to a more collective and organization perspective.”

Consumer vs. Citizen Rights Framework

Another critical consequence identified was the reduction of people from citizens with rights to consumers in a marketplace. At the Innovative Regulatory Strategies session, Dr. Gillian Marcelle observed that “people are being treated like consumers in a neoliberal framing rather than citizens who have the right to communication and the rights to access.”

Balancing Competing Rights

The challenge of balancing individual privacy rights with other human rights was discussed at the Advancing Data Governance session. Joseph from Wikimedia Foundation asked about “how, through this process of harmonizing regional data protection laws and implementing such new laws, how we can ensure that all human rights are respected throughout this process and that the right to privacy does not come at the expense of any other potential right.”

Digital Solidarity vs. Individual Sovereignty

The concept of moving beyond individual digital rights toward collective approaches was further explored through discussions of digital solidarity. At the AI at a Crossroads session, Ana Valdivia advocated for “digital solidarity” rather than individual digital sovereignty, stating “rather than talk about digital sovereignty, that creates sort of like frictions between states, because all the states in the world want to become digital sovereign, we should talk about digital solidarity.”

Expanding to Economic and Environmental Rights

The discussions also highlighted the need to move beyond first-generation civil and political rights to include economic and environmental considerations. Allison Gilwald emphasized “extending that if one’s really concerned about redressing the current inequalities we see from a rights perspective is shifting that to second- and third-generation rights. So to economic and environmental rights.”

No relevant discussions found.

The question of enhancing data collection efforts to better capture diversity among persons with disabilities was discussed across several sessions at the Internet Governance Forum 2025, with the most comprehensive insights provided in disability-focused sessions.

In WS #69 Beyond Tokenism Disability Inclusive Leadership in IG, Dr. Derrick Cogburn provided the most detailed framework for disability data collection, highlighting existing resources: “we have two really, really good sources of disability data. One is called the Disability Data Initiative, which is led by Fordham University, and the Disability Data Hub, led by the World Bank. Both of these data sets, as well as the text data, provides tremendous data for us to be able to analyze how persons with disabilities are faring in this current period.” He emphasized that “most of this data is open data” and stressed the importance of “continuous capacity development in research capacity” to effectively utilize these tools.

A practical example of inclusive data collection was shared in Open Forum #21 Leveraging Citizen Data for Inclusive Digital Governance, where Omar Seidu from Ghana Statistical Service described their comprehensive approach: “And understanding the key stakeholders’ interest for engagement. That is very important because once you engage, in fact, in developing this, there are 12 different domains of persons with disabilities in Ghana. And we make sure we engage with all these different domains. And that helps us to develop the functionalities for persons who are partially blind, for persons who have hearing impairment and all that, to engage with the application.”

The importance of disability status as a segmentation factor in data collection was acknowledged in Open Forum #29 Advancing Digital Inclusion Through Segmented Monitoring, where Pria Chetty noted: “And then, of course, in our data, we have that valuable demographic segmentation. So that’s by gender, age, income level, education, location. Anika mentioned to include the peri-urban category in there, but also disability status and language as well.”

In Open Forum #82 Catalyzing Equitable AI Impact, the need for inclusive representation was highlighted, with Nupur Chunchunwala emphasizing that “humans are diverse. We have an aging population over 10 percent that’s going to get impacted. We have, of course, gender. We have ability in terms of disabilities that are coming on and a large population of neurodiverse individuals.”

The discussions revealed that effective data collection for persons with disabilities requires comprehensive stakeholder engagement across different disability domains, the utilization of existing open data sources, investment in research capacity development, and the integration of accessibility features in data collection platforms to ensure meaningful participation of all disability communities.

Based on discussions across multiple IGF 2025 sessions, several key directions emerge for national, regional, and global efforts regarding taxation in the digital economy.

Regional Tax Optimization Initiatives

The African Union is leading significant efforts in digital taxation reform. At the African Union Open Forum, Maktar Sek from UNECA revealed comprehensive regional work: “Why we have developed one platform, tax calculator, to review the taxation in the ICT sector. I think we can share the link at the screen and this taxation has been conducted in 54 member state.” The focus is on optimization rather than revenue maximization, with Sek emphasizing that “Optimizing ITC tax can increase not only the GDP, but we have seen an increase according to our statistic of the broadband connectivity as well as the job creation.”

Device Taxation Reform for Digital Inclusion

A critical area for national efforts involves reducing taxation on digital devices to improve affordability. At the Connectivity Gap workshop, Onica Makwakwa highlighted the impact of device taxation: “Looking at affordability specifically one of the things that we’ve done successfully in a couple of countries is looking at taxation of devices and found we found that there’s anywhere from 20 to 40 percent of Taxation that is on devices whether it’s an import duty tax your VAT Or you know sales tax or what have you but there’s anywhere from 20 to 45 percent And we’ve been able to demonstrate actually that if governments could just roll back some of those taxes It actually increases uptake of digital technologies within the country.”

Digital Services Taxes and Platform Taxation

Multiple sessions addressed emerging models for taxing digital platforms. At the Editorial Media session, Anya Schiffrin outlined recent developments: “Late 2024, Australia announced that platforms that didn’t want to negotiate with publishers could pay a digital levy, which would be more expensive than what bargaining code payments would be. South Africa is also looking at digital levies. And many countries are considering things like digital services taxes with funding earmarked for journalism.”

Cross-Border Fiscal Cooperation

Global coordination efforts are gaining momentum. At the WSIS+20 dialogue, Juan Carlos Lara emphasized the need for international fiscal measures: “Much of this work requires funding, sustainable digital cooperation requires public investment and ad hoc funding may not do that or provide that in a sustainable way. So to link the financing for development process, to acknowledge the need for cross-border fiscal measures is also relevant here, including equitable taxation of digital services and other ways to generate resources for digital infrastructure.”

Digital Development Tax Proposals

New concepts for addressing digital inequality through taxation are emerging. At the AI Governance session, William Bird mentioned recent proposals: “The Global Digital Justice Forum, in fact, just made some of its submission to the WSIS Plus 20, and one of the things that they’re calling for is a digital development tax that should be imposed on these entities in order to fundamentally address this inequality.”

International Coordination Challenges

Future efforts face significant geopolitical challenges. Schiffrin raised critical questions about international cooperation at the Editorial Media session: “Can Europe stick together? In the US, those of us who care about this stuff want to know whether the EU and the rest of the world will cave, capitulate, or whether it will stick with its plans to tax and regulate big tech.”

Based on the meeting transcripts analyzed, the question of finding a middle-ground solution between advancing global digital trade agreements and addressing immediate challenges like digital divides and data fairness was only partially discussed in one session.

In WS #259 Multistakeholder Cooperation in Era of Increased Protectionism, speakers touched on both aspects of this challenge. Milton Mueller raised concerns about digital free trade and data movement across borders, specifically questioning WTO negotiations and multistakeholder participation. Flavia Alves highlighted the complexity of balancing different priorities, stating that “data flows is also not only an economic issue, but it’s also a privacy and security issue, that we need to be careful and balance how we do the safeguards on privacy at the same time on law enforcement and others.”

Regarding the more immediate challenges, Tatjana Trupina emphasized the urgency of addressing basic connectivity issues, noting that “the original WSIS goal was connectivity, we still have one third of the world not connected, the WSIS has to deliver, we have to strengthen its implementation to address the current and emerging digital divides.”

While this session provided some insights into the tension between advancing digital trade and addressing fundamental digital divides, no concrete middle-ground solutions were explicitly proposed or discussed in detail across the analyzed sessions.

Creating Meaningful Accountability Mechanisms for Big Tech Companies Beyond Fines

The discussions across multiple sessions at the Internet Governance Forum 2025 revealed widespread consensus that traditional fines alone are insufficient to drive meaningful changes in big tech corporate behavior. Speakers emphasized the need for more innovative and comprehensive accountability mechanisms.

Structural and Regulatory Approaches

Several sessions highlighted the need for structural changes beyond monetary penalties. In the competition rights workshop, Camila advocated for “bolder theories of harm” and suggested “breaking up companies” as a solution, stating “as we see that they have an unmeasurable impact in our lives, maybe the solution is that they didn’t have to be that big.”

The child safety high-level session emphasized that regulation combined with enforcement works, with Thibaut Kleiner noting “You cannot just count on the goodwill of companies that are making profits to change their features unless they have really some pressure also coming from the regulators.”

Transparency and Due Diligence Requirements

Multiple sessions emphasized transparency as a key accountability mechanism. In the AI policy forum, Wai Sit Si Thou advocated for “a public disclosure accountability mechanism that could reference the ESG reporting framework” with “public disclosure on how this AI works and its potential impact.”

The human rights session highlighted the need for mandatory human rights due diligence, with Peggy Hicks stating “we do need to to create the right types of both incentives and disincentives For companies to actually, you know Do the risk assessment that needs to happen.”

Innovation in Financial Accountability

Beyond traditional fines, speakers proposed creative financial mechanisms. In the media dependency session, Anya Schiffrin suggested requiring platforms to “post a billion dollar bond before they start operations in countries like Kenya or Brazil, so that when there’s a fine, they have to pay it.”

The main session discussed social offset mechanisms similar to carbon offsets as an innovative accountability approach.

Dialogic and Collaborative Approaches

The German regulator session presented a “dialogic regulation” model where Michael Terhorst explained: “we don’t just, yeah, send a letter saying, okay, you have to, I don’t know, pay €5 million because your platform isn’t safe. We get in touch with the provider and say, okay, we found some deficits on your platform.”

The parliamentary session on vulnerable groups advocated for collaborative design, with Neema Iyer suggesting “what would be lovely in a really perfect world would be if these algorithmic decisions are co-created by all of us.”

Legislative and Parliamentary Engagement

Parliamentary sessions emphasized direct engagement with tech executives. In the cybercrime parliamentary session, Pavel Popescu advocated for “bringing platform CEOs before parliamentary committees” to ask them “very complicated and direct questions.”

International Cooperation and Collective Action

Multiple sessions emphasized the need for coordinated global action. The African cybercrime forum highlighted regional coordination, with Senator Besalisu advocating that “unless and until all of those big texts know that if you violate the law in South Africa, your sanction is not limited to South Africa.”

The parliamentary cybercrime session discussed creating “for big tech consequences at the global level, at the UN level” including potential binding treaties.

Investor and Market-Based Mechanisms

The conflict accountability session explored investor pressure as an accountability tool, with Kiran Aziz explaining how investors can exclude companies from portfolios, providing “a quite thorough exclusion document and this is a way to hold companies accountable but this is also to help other investors to get insight about where we draw the line.”

Public Procurement as Leverage

The cloud autonomy session highlighted public procurement as a powerful tool, with Agustina Brizio noting that “public procurement is amazing level here” and advocating for governments to include “in every public procurement contract things as data localization, as having over standards, as having a more transparent governance.”

Civil Society and Multi-Stakeholder Oversight

The digital rights partnership session emphasized civil society’s watchdog role, with Ian Barber noting that “civil society can play a key role…serving as kind of a watchdog or an observer” and that “a lot of this comes down to transparency, openness, and decision-making in the processes.”

Key Challenges and Gaps

Despite these various approaches, significant challenges remain. The <a href="https://dig.watch/event/internet-governance-forum-2025/day-0-event-255

No relevant discussions found.

Risks of Over-Reliance on Self-Regulation and Corporate Social Responsibility

The discussions across multiple IGF 2025 sessions revealed significant concerns about relying too heavily on self-regulation and corporate social responsibility to address tech-related societal challenges. A clear consensus emerged that self-regulation has largely failed across various domains, from child safety to content moderation to platform accountability.

Evidence of Self-Regulation Failures

Multiple speakers provided concrete evidence of self-regulation failures. In High Level Session 4, Leanda Barrington-Leach definitively stated that “Self-regulation has not worked and good regulation does work.” She further emphasized that “whistleblower reports and leaked internal documents show how time and again tech companies are aware of the harm they are causing children and choosing to do it anyway.”

In Main Session 2, Jhalak Kakkar provided historical context, noting that “we’ve also seen that companies have never been particularly adept at only working under the realm of self-regulation. I mean, whether, and this is across industries, I’m not only pointing to tech, you know, we’ve seen that time and time again over the last 150 years.”

Specific Examples of Self-Regulation Inadequacies

Several sessions provided specific examples of self-regulation failures. In Day 0 Event #255, Marwa Fatafta criticized inadequate corporate self-assessments, explaining how “Microsoft had recently issued a statement after a year and a half of public mobilization… in which they said, well, we conducted an audit to see whether our technologies have contributed to harm or targeting of civilians in the Gaza Strip. And while we don’t have an insight into how our technologies are used by Israel, especially in air-gapped military bases, we concluded that we have not contributed to any harm.”

In Global Youth Summit, Brendan Dowling explained Australia’s rationale for government intervention, stating that “the response of social media companies has been lacklustre at best, disingenuous at worst.”

Structural Problems with Self-Regulation

Several speakers identified fundamental structural problems with self-regulation. In Lightning Talk #109, Michael Terhorst highlighted the objectivity problem, noting that “when we look at some voices saying that there is already enough. We hear from some people that when we look at social media platforms for example that providers already do a lot to protect the personal integrity of minors. Those voices are from the providers themselves. So, maybe it is not really objective.”

In NRI Collaborative Session, Dennis Broeders explained the fundamental conflict of interest: “I think the general rule is when industries start saying, no, we’ll do it ourselves, we’ll self-regulate, we’ll do ethical framework, we’ll do all these things, that means they’re afraid of regulation, right, that’s the only reason they’re stepping forward to do this.”

Power Imbalances and Resource Disparities

A critical concern raised was the significant power imbalance between tech companies and regulatory bodies. In Open Forum #48, Leanda Barrington Leach highlighted that “There is very little knowledge, there is very very little capacity and as I mentioned before there is a huge amount of resources on the other side.” She emphasized there is “a massive massive imbalance of power and resources between in this case the tech sector.”

Solutions and Approaches

Multiple sessions discussed solutions to address the failures of self-regulation. In Main Session 2, Jhalak Kakkar emphasized the need for “external regulators, we need communities to be engaging, a bottom-up approach, civil society to be engaging, multilateral institutions to be coming in.”

In High Level Session 1, Thibaut Bruttin argued that “Governments should not be afraid of regulating and legislating” and that their “responsibility is to build the framework, not to go into the nitty-gritty details of everything happening in the media field, but really building this framework that enables media to flourish.”

In Day 0 Event #252, Tawfik Jelassi concluded that “Self-regulation did not work, did not deliver. We need maybe a core regulatory system, which is truly multi-stakeholder.”

Balanced Approaches

Some speakers advocated for balanced approaches that combine regulation with incentives for responsible corporate behavior. In WS #106, Julija Kalpokiene noted that “if the right incentives exist, then self-regulation could be much better because the industry know what they’re able to do, so it can be more effective and better focused than the government approach.”

However, the overwhelming consensus across sessions was that effective regulation with enforcement mechanisms is essential to address the societal challenges posed by technology platforms and companies, as self-regulation alone has proven inadequate to protect public interests.

No relevant discussions found.

No relevant discussions found.

No relevant discussions found.

The operationalization of international norms on cybersecurity and critical infrastructure protection was extensively discussed across multiple sessions at IGF 2025, revealing both established frameworks and significant implementation challenges.

Existing Framework and Challenges

Several speakers emphasized that while international norms exist, their real-world implementation remains problematic. As discussed in the Open Forum on Critical Infrastructure, Pavel Mraz explained that “in order for the UN framework to have a real world impact and not remain just on paper, it must be operationalized nationally through legislation, institutional coordination, but also sustained investment in cybersecurity”.

The gap between established norms and practice was highlighted in the ransomware accountability session, where Julie Rodriguez Acosta noted that “while the framework says the reality on the ground tells us a different story”.

Multi-Level Coordination Requirements

Speakers consistently emphasized the need for coordination across multiple levels. Caroline Troein, in the critical infrastructure forum, stressed that coordination “needs to happen at the national regional and global levels”.

This multi-level approach was reinforced in the subsea cables protection session, where Minister Tung stated: “We need a combination of national, regional, and international cooperation to achieve effective resilience measures”.

Practical Implementation Mechanisms

Several concrete operationalization mechanisms were discussed:

Legal and Regulatory Frameworks

Giacomo Paoli Persi suggested in the ransomware session developing “model law or a model legislation for those countries that need to adopt some sort of regulatory measures”.

Crisis Communication Mechanisms

Pavel Mraz highlighted the importance of establishing crisis communication channels, noting in the critical infrastructure session that “countries are designating points of contacts globally for crisis communication in recognition that you cannot exchange business cards in a hurricane”.

Regional Cooperation Examples

Practical examples were shared, including Under-Secretary Syrjala’s mention in the subsea cables session of “the recent Memorandum of Understanding, which the Baltic Sea, NATO Allies and the EU have published”.

Implementation Challenges

Despite established frameworks, significant challenges remain. In the conflict and crises session, Dennis Broeders noted that “Implementation is really hard when it comes to negative norms”.

The Dynamic Coalition session highlighted that despite standards development efforts, “once these standards are developed not very many governments or even institutions non-government institutions even use them”.

Integration with International Law

The need for legal integration was emphasized in the Digital Emblem session, where Samit D’Chuna explained that “there need to be these common understandings of what the digital emblem is and how it’s respected and what happens when it’s not respected”.

The discussions revealed that while comprehensive international frameworks exist, successful operationalization requires sustained commitment across technical, legal, diplomatic, and institutional dimensions, with particular emphasis on bridging the gap between high-level agreements and practical implementation at national and regional levels.

The global race for quantum supremacy and its geopolitical and economic implications received limited discussion across the Internet Governance Forum 2025 sessions. While most sessions did not address this topic directly, some relevant observations emerged from discussions on related technologies and cybersecurity.

The most substantive mention came from the Dynamic Coalition Collaborative Session, where Elif Kiesow Cortez highlighted the divergent national approaches to post-quantum cryptography, noting “US and EU, they have some distinct but also converging approaches” and referenced specific national programs including “France through ANSI also advocating for hybrid solutions, or Germany through BSI providing guidance” as well as developments “in the Netherlands as well, we have a PQC migration handbook.”

Additionally, quantum computing was briefly referenced in the context of cybersecurity threats during WS #193 Cybersecurity Odyssey Securing Digital Sovereignty Trust, specifically regarding the need for post-quantum cryptography preparations.

The discussions suggest that while the broader implications of quantum supremacy were not extensively explored, there is growing recognition among different nations and regions of the need to prepare for quantum computing’s impact on cybersecurity infrastructure, with various countries developing distinct but sometimes converging approaches to post-quantum cryptographic standards.

The question of ensuring quantum computing developments are accessible and beneficial beyond a handful of leading countries or corporations received limited attention across the IGF 2025 sessions reviewed. Only two sessions touched upon this critical issue, with minimal substantive discussion.

The most relevant discussion occurred during the Launch / Award Event #169 Report Launch: Quantum encryption: blessing or havoc?, where Wout de Natris briefly addressed the global equity concerns, stating that “the whole world starts acting, that nobody’s left behind, that developing nations are assisted to set the steps they need to take. Because if they are vulnerable, we are vulnerable.” However, this discussion remained brief and did not delve into specific mechanisms or strategies for ensuring equitable access.

The Dynamic Coalition Collaborative Session provided a tangential perspective through Elif Kiesow Cortez’s recommendations emphasizing “creating global standards that look at interoperability” to ensure “we are not leaving any organizations, countries, people behind when it comes to making our more secure movement towards the better Internet.” While this touched on inclusivity principles, it did not specifically address quantum computing accessibility.

Overall, the question of quantum computing equity and global accessibility remains largely unaddressed in the IGF 2025 discussions, representing a significant gap in the dialogue around emerging technologies and digital inclusion.

The ethical, legal, and societal implications of emerging neurotechnologies received minimal attention across the Internet Governance Forum 2025 sessions, with only brief mentions in two workshops.

The most substantive reference came from Open Forum #82, where UNESCO’s Mariagrazia Squicciarini highlighted the organization’s recent work on neurotechnologies. She noted that “The latest recommendation that has been worked on is about neurotechnologies and the impact they have on rights, on the people, on, again, what society wants them to do or not to do.” She emphasized particular concern about “the crossroad between AI and neurotech, because that’s where the biggest impact may be on societies.”

A related concern about mental privacy emerged in Workshop #395, where Nieves Molina touched on the evolving nature of freedom of thought in the digital age. She observed that “the issue of freedom of thought now takes a different connotation because technology actually my advance to that point that you that what you think can be reachable even if you don’t announce it as technology will advance and we will have to take decisions on whether the legislation that we have covers.”

Despite these brief acknowledgments of the importance of neurotechnology governance and mental privacy protection, there was no comprehensive discussion about the need for new frameworks for mental privacy and cognitive freedom across the forum sessions.

Biotechnology and Neurotechnology: Blurring Lines Between Humans and Machines

The question of how advances in biotechnology and neurotechnology might blur the line between humans and machines, and what new rights or protections might be needed, received limited but notable attention across the Internet Governance Forum sessions.

Brain-Computer Interfaces and Human Enhancement

The most direct discussion occurred in High Level Session 3: AI & the Future of Work, where Ishita Barua highlighted the transformative potential of brain-computer interfaces. She noted: “And perhaps most astonishing, brain-computer interface in combination with AI. With the help of these devices, people with paralysis can regain the ability to move, speak, and even write through direct decoding of brain activity.” However, the session did not delve into the broader implications for human-machine boundaries or necessary protections.

International Governance and Rights Framework

A more comprehensive approach was discussed in Open Forum #82: Catalyzing Equitable AI Impact, where UNESCO’s work on neurotechnology governance was highlighted. Mariagrazia Squicciarini explained: “The latest recommendation that has been worked on is about neurotechnologies and the impact they have on rights, on the people, on, again, what society wants them to do or not to do. And the special attention is also put at the crossroad between AI and neurotech, because that’s where the biggest impact may be on societies.”

Gap in Discussion

Despite the significance of this topic for future internet governance and human rights, the majority of IGF sessions did not address the convergence of biotechnology, neurotechnology, and digital systems. This suggests a need for more focused attention on these emerging technologies and their implications for human identity, rights, and protections in future governance discussions.

The discussions across various IGF 2025 sessions revealed limited but emerging attention to virtual reality and metaverse technologies, with most focus on potential applications rather than comprehensive risk assessment.

Educational Applications and Opportunities

Several sessions highlighted the transformative potential of VR in education. In the Open Forum on DPI and Open Source AI, Judith Okonkwo discussed practical implementations: “VR for Schools initiative that we have, and this is now looking at deploying this technology in really resource-constrained learning environments, so how can we go into a situation where, for example, you have a school without, you know, infrastructure for things like science experiments, right, and that kind of resource-constrained environment, can you bridge the the gap with immersive tools, right, can you create a VR lab where students are then able to do simulations.”

The Lightning Talk on Fundamental Rights in Metaverse presented compelling statistics, with Avinash Dadhich noting that “research says that by 2026, more than 70% of students in UK, they will be in the virtual world, in the 3D world for at least one hour in a day” and predicting that “maybe next 10, 15, 20 years we will be a museum of mobiles, and then there will be a virtual world where people will be interacting.”

Social Interaction and Accessibility

The discussions revealed interesting applications for social interaction, particularly for marginalized groups. The Creative Workshop on Tech-Driven Solutions explored how “metaverse if if if it’s completely online we would have more control let’s say if it’s online to do the payments of some of some digital transactions and some people are very visual let’s say which may be not not so into social interactions but very visual so metaverse could help online or it could be augmented reality” could assist people with disabilities.

Indigenous communities are also exploring these technologies. In the Open Forum on Indigenous Peoples’ Languages, Outi Kaarina Laiti discussed “developing content in platforms like Second Life, Minecraft and so on, which I call the indigenous metaverse. It’s growing rapidly and most of these platforms are private-owned.”

Risks of Exclusion and Digital Divides

The risk of exclusion was prominently addressed in the Dynamic Coalition Collaborative Session, where Jutta Croll asked: “Do you think we are at risk of having such a, like a second digital divide in regard of emerging technologies?” Dino Cataldo Dell’Accio confirmed this risk exists, highlighting the concept of a ‘metaverse divide’ and emphasizing the need for ‘inclusiveness by design’ in emerging technologies.

Manipulation and Privacy Concerns

Concerns about manipulation were raised in the Lightning Talk on Fundamental Rights, where Anurag Vijay illustrated how “I was speaking to my friend over phone about a Rado watch… The next day, ads were flashing all over my social media” and questioned whether “Do you think we need to be protected in that way? Because our decision is being influenced by the artificial intelligence.”

Governance and Child Protection

Child protection in metaverse environments was addressed in the IGF Intersessional Work Session, where Jutta Croll mentioned that current issues include “child rights, recognition while shaping and developing the metaverse, or the engagement in developments of privacy and anonymity preserving mechanisms of age verification solutions as a precondition for a secure and safer usage of the digital environment.”

The broader governance challenge was acknowledged in High Level Session 5, where Fabrizia Benini from the European Commission questioned whether current governance models would remain robust “facing, in face, of the new technologies? Will the governance model hold?”

Overall, while the discussions showed growing recognition of VR and metaverse potential, there was limited comprehensive analysis of labor market impacts or detailed strategies for addressing addiction risks, indicating areas requiring further attention in future governance discussions.

The question of metaverse governance and content moderation rules was primarily addressed in the Lightning Talk #143 Fundamental Rights in Metaverse session, where several governance models were explored.

Utkarsh Leo outlined different approaches to metaverse governance, explaining the centralized model: “if you look at a more centralized-based model wherein there is one company, let’s just for the sake of saying, let’s say if meta has all control over the metaverse, then in a way we’re leading towards a more monopolistic form of a world where perhaps suppression can happen.”

As an alternative, he presented the decentralized approach: “Other proposals say that let’s have a very decentralized world which focuses on blockchain technology and a distributed consensus-based model.”

The discussion concluded with a proposed hybrid solution: “a hybrid model wherein you may have a central entity, but at the same time certain rights are recognized through blockchain and distributed consensus-based mechanisms.”

This topic was not substantively discussed in other IGF 2025 sessions, indicating that metaverse governance remains an emerging area requiring further attention in internet governance forums.

The Role of International Institutions in Setting Norms for Advanced Technologies

The discussions across IGF 2025 sessions revealed a strong consensus that international institutions must play a central role in governing advanced technologies that challenge traditional regulatory frameworks, while emphasizing multi-stakeholder approaches and inclusive participation from the Global South.

The Imperative for International Cooperation

Multiple speakers emphasized that technological challenges transcend national borders and require coordinated international responses. As noted in the High Level Session on AI & the Future of Work, “I’m very sure that we address these challenges best with global cooperation. I don’t think that every single nation can make up some kind of framework that will protect people from any kind of risk. We have to find international regulations.” This sentiment was echoed in the Open Forum on Shaping Global AI Governance, where Ambassador Noorman stated that “no country can solve this alone. All AI transcends borders. So must be our response.”

UN Leadership and Global Frameworks

The United Nations emerged as a critical institution for establishing global norms. In the Open Forum on Building an International AI Cooperation Ecosystem, Wolfgang Klauerwachter argued that “If it’s a global problem, we need all countries on the table, and only the United Nations offers this opportunity.” The Opening Ceremony highlighted key UN initiatives, with Li Junhua noting negotiations for establishing “the Independent International Scientific Panel on Artificial Intelligence and the Global Dialogue on AI Governance within the United Nations.”

The Global Digital Compact was repeatedly cited as a foundational framework. In Main Session 2, the panel discussed how this compact represents “the first time every UN member state in the world came together to agree a path on AI governance.”

Regional and Specialized International Bodies

Regional institutions also play crucial roles in norm-setting. The Council of Europe’s work on AI was highlighted in multiple sessions. In the High Level Session on Information Space, Bjorn Berge explained that “the Council of Europe has now agreed and concluded a new international treaty on Artificial Intelligence and Human Rights, Democracy and Rule of Law. The first international treaty of its kind.”

UNESCO’s role in setting ethical standards was emphasized across multiple sessions. In the Workshop on AI Innovation, Guilherme Canela de Souza Godoi noted that “more than 70 UNESCO member states have already implemented the readiness assessment methodology.”

Cybersecurity and Critical Infrastructure Norms

For cybersecurity governance, the Open Forum on Cyber Resilience highlighted the UN framework for responsible state behavior. Pavel Mraz explained that “The UN framework for responsible state behavior in cyberspace… does provide a strong foundation for protecting critical infrastructure.”

Multi-Stakeholder Governance Model

A consistent theme was the need for multi-stakeholder approaches rather than purely state-centric processes. The Workshop on WSIS+20 raised concerns about “risks of more state-centric process to the appointment of experts, exclusion of military applications from the scope of the assessments” in AI governance mechanisms.

Challenges and Limitations

Several speakers acknowledged significant challenges in international norm-setting. In the Workshop on Bridging Internet AI Governance, Bill Drake expressed skepticism, noting that “multilateral regulatory interventions are impossible to contemplate in the Trump era” and questioned what kind of binding international agreements could realistically be negotiated.

The complexity of coordination was highlighted in the High Level Session 5, where Thomas Schneider noted that “you will have a distributed system with thousands of actors in the end involved in actions on global regional and national levels and you can’t put them all you can’t even put them all in one in one room”.

Global South Inclusion and Equity

A critical concern raised throughout the discussions was ensuring Global South participation in norm-setting processes. In the Open Forum on High Level Review of AI Governance, Abhishek Singh emphasized “we need to bring countries of Global South at the decision-making tables.”

Interoperability Over Uniformity

Rather than seeking uniform global governance structures, many speakers advocated for interoperability between different regulatory approaches. In the Workshop on AI Policy Research, Anne Flanagan argued that “We’re never going to have a global regime. We’re never going to have a single global governance structure. It’s not realistic. It’s also not appropriate” because different regions have different cultural contexts and priorities.

Implementation and Enforcement Challenges

The gap between norm-setting and implementation was a recurring concern. In the Open Forum on Autonomous Weapon Systems, Chris Painter noted the challenge that “There’s now a dispute about what those norms mean, that’s what international law means, and the biggest part… is how do you have accountability once there’s agreement among countries.”

The discussions revealed a complex landscape where international institutions are seen as essential for addressing the global nature of advanced technologies, while acknowledging significant challenges in achieving consensus, ensuring inclusive participation, and implementing effective governance frameworks that balance innovation with protection of rights and security.