Data & Indicator Lab: Multistakeholder Workshop on WSIS Action Lines, Indicators, and Accountability

Nils Berglund argues that WSIS has accumulated broad shared commitments over two decades, but there is still no dedicated measurement system tied directly to the WSIS process. As a result, it is difficult to compare progress over time or across countries and to hold states and stakeholders accountable for what they have committed to.

He says that despite 20 years of converging commitments and a broad normative consensus on what the information society should look like, implementation remains fragmented and inconsistent partly because there has been no real framework for measuring progress directly tied to WSIS . He adds that this makes it hard to say with comparable data whether countries are doing better or worse than five years ago and therefore harder to maintain accountability, while the WSIS stocktaking platform functions mainly as a best-practice repository rather than true accountability infrastructure .

Nils Berglund argues that the current WSIS review moment creates an opening for a broader range of stakeholders to contribute directly to monitoring and evaluation. His point is that civil society, academia and the technical community already produce relevant data, and this should be brought into the WSIS process to strengthen monitoring infrastructure.

He highlights that beyond UN datasets there is a large ecosystem of civil society, academic and technical community measurement, and says this is a good opportunity to discuss what the UN can learn from those sources . He later says the current WSIS Plus 20 implementation architecture creates an opportunity for everyone in the room and the broader multistakeholder community to contribute data, strengthen indicators and help build better monitoring infrastructure .

Nils Berglund argues that headline improvements in connectivity should not be treated as sufficient evidence of progress because they can obscure serious negative outcomes in rights and freedom. He is urging a more outcome-based approach that asks what kind of digital society WSIS is actually trying to achieve, not just how many people are nominally connected.

He cites civil society data showing that in 2025 there were 313 shutdowns in 52 countries, with the number increasing year on year, which indicates a negative trend that is not always visible in global datasets . He also notes that although 5.5 billion people are connected to the Internet, an estimated 81 percent of them live in countries where individuals were arrested or imprisoned for posting on political, social or religious issues, demonstrating that connectivity statistics alone can give the wrong picture .

Nils Berglund argues that there is already a substantial amount of relevant ICT data from both UN and non-UN sources, but its coverage is uneven across policy areas. In his view, measurement is strongest on infrastructure and access, while rights, media, ethics and other difficult governance areas remain much more weakly covered.

He points to established UN data sources such as the ITU's ICT Development Index, UNDESA's e-government and e-participation indices and UNCTAD's work on the digital economy as evidence that there is broad coverage on paper . He then says the data are uneven because connectivity and infrastructure data are comparatively mature, while data on rights, media and ethics are much thinner, and many countries still do not provide gender-disaggregated data .

Esperanza Magpantay explains that the Partnership on Measuring ICT for Development has a formal UN-backed mandate to review WSIS monitoring indicators, methodologies and data availability. She presents this as a structured, multi-stakeholder process that will gather feedback and report to the CSTD in 2027.

She explains that the partnership was created out of WSIS, exists to improve the availability, quality and international comparability of information society indicators, and coordinates the work of 14 organisations through a common core list of indicators endorsed by the UN Statistical Commission . She then states that the 2025 UN General Assembly WSIS resolution mandated the partnership to review indicators covering WSIS action lines and the SDGs, to work with stakeholders in a systematic review, and to report findings to the 2027 session of the CSTD, including feedback collected through a stakeholder form .

Esperanza Magpantay argues that while the review should be open to new stakeholder proposals and broader coverage, the partnership must preserve international comparability. Her position is that revised indicators are only useful for the partnership's purposes if they can be applied consistently across countries.

In her closing response, she welcomes the suggestions and asks participants to submit them in writing so they can be considered in the review . She adds that international comparability is one of the partnership's main objectives and asks contributors to take this into account when proposing new indicators or methods .

Anriette Esterhuysen argues that the standard Internet penetration measure is too crude because it treats minimal use as equivalent to substantive connectivity. This obscures the difference between people who are fully and meaningfully connected and those who only connect rarely, producing misleading conclusions about digital inclusion.

She specifically criticises the indicator that counts someone as connected if they access the Internet once in a three-month period, saying this masks the gap between the hyper-connected and the barely connected . She links this to the growing recognition of meaningful connectivity through the ITU Broadband Commission and says the current measure distorts the relationship between demand-side and supply-side data .

Anriette Esterhuysen argues that better demand-side data are essential, especially for people working with governments in developing countries. She warns that current connectivity statistics can be misused to exaggerate success and conceal the continuing scale of exclusion.

She welcomes the more demand-side-oriented initiative and says such data are extremely important for those working on the ground in developing countries and with governments . She also warns that data can be abused and says connectivity data are currently being used to inflate progress, disrupt understanding and hide the huge remaining challenge of connecting the unconnected .

Pria Chetty argues that measurement should move beyond counting infrastructure and access and instead focus on digital inclusion, actual user experience, safety and sectoral outcomes. Her point is that indicators need to capture how connectivity is experienced and what it enables in practice across areas such as health, education and agriculture.

She describes Research ICT Africa's long-running household, microenterprise and gender-focused surveys and explains that the 'After Access' approach was designed to move beyond measuring connectivity alone and towards digital inclusion . She then says measurement must move away from infrastructure discussions to understand experience, safety and sectoral realities, and that indicators should align with strategic priorities in health, education and agriculture .

Pria Chetty argues that quantitative indicators alone are not enough and need qualitative elements to become meaningful and policy-relevant. She sees this as necessary to make the data understandable, accessible and useful for shaping digital transformation and inclusion strategies.

She says there is an increasing call to build qualitative elements into indicators so they provide meaning and can be used more effectively in decision-making . She gives examples by asking how indicators can be structured to explain strategic priorities and dependencies in health, education, agriculture, digital transformation, skilling and digital inclusion .

Pria Chetty argues that national statistical offices need much closer working relationships with regulators and local government if data are to be usable and actionable. Without these links, data cannot be properly accessed, translated or triangulated at the levels where policy decisions are made.

She identifies a recurring barrier in getting statistical offices to build the necessary data infrastructure and says they need stronger relationships with information regulators so indicator work becomes accessible and translatable where needed . She also argues that statistical offices should work more closely with local government to reach citizens and baseline data, warning that otherwise civil society faces an unnatural burden to represent citizen realities without adequate cooperation from local authorities .

Pria Chetty argues that combining spatial data with gender data exposes patterns of exclusion that standard aggregate indicators miss. This produces a more realistic picture of barriers affecting skills, jobs, digital inclusion and the uptake of AI.

She refers to newly released gender-focused studies and says it becomes very revealing when spatial data are combined with gender data . She explains that this combination uncovers new traps and barriers affecting skilling, digital inclusion, job trajectories and how AI might be absorbed, leading to a more informed picture .

Bridgette Hanani Ndlovu argues that infrastructure and connectivity metrics on their own do not show whether people can actually exercise digital rights in practice. She stresses that countries may have the infrastructure in place while simultaneously restricting access through shutdowns and other abuses.

She says that from an African perspective there is a need to consider interlinkages with human rights mechanisms, and cites Paradigm Initiative's research assessing state compliance with the African Commission on Human and People's Rights Declaration on Principles of Freedom of Expression and Access to Information in Africa . She notes that many countries have infrastructure that generates positive quantitative data, yet there are major challenges in whether that infrastructure is used appropriately, including frequent and prolonged Internet shutdowns that prevent meaningful use by communities .

Bridgette Hanani Ndlovu argues that WSIS monitoring should explicitly include human rights dimensions and not treat infrastructure deployment as sufficient evidence of progress. Her point is that assessing respect for freedom of expression and access to information is necessary to understand whether connectivity actually benefits people.

She explicitly asks how issues around rights will be integrated into the assessments and says there is a need to connect measurement to specific human rights mechanisms . As an example, she points to research on compliance with the African Commission's declaration on freedom of expression and access to information, and contrasts this with evidence of shutdowns in countries that otherwise appear well covered by infrastructure indicators .

Bridgette Hanani Ndlovu argues that access indicators should account for affordability and not only formal availability or policy intent. She shows that governments may claim to support device access while taxation and pricing make real uptake impossible for many people.

She gives the example of Malawi, where there is a policy aimed at improving access to digital devices . She says that heavy taxation on those devices means affordability remains a serious problem, so indicators may report high access figures even though many people cannot actually afford digital technologies on the ground .

Nandini Chami argues that gender is being structurally overlooked in WSIS-GDC implementation tracking. She warns that in practice, gender mainstreaming has too often meant gender disappearing from the framework rather than being integrated seriously.

She says the draft WSIS-GDC implementation roadmap submitted to the CSTD in April mentions the word 'gender' zero times . She links this omission to the lack of a dedicated action line on gender and states that mainstreaming gender must not become a way of 'streaming gender away' .

Nandini Chami argues that the problem is not a total absence of gender-related data, because many useful indicators already exist across UN agencies and related bodies. She is calling for these existing gender-disaggregated sources to be incorporated directly into WSIS action line monitoring.

She says there are already gender indicators collected by WSIS action line facilitating agencies beyond simply measuring the gender divide in access . As examples, she lists WIPO's patent scope analysis on gender-disaggregated patent inventions, the ILO digital labour policy tracker, national time-use, income and employment statistics, the UN Women global database on violence against women, and the WHO global digital health monitor indicator, arguing that these could be directly used in action line tracking .

Nandini Chami argues that the rise of AI changes the measurement challenge and requires deliberate investment in new indicators. She stresses that the goal should not be merely integrating women into ICT markets, but tracking whether AI-driven structural transformation is creating meaningful pathways to empowerment.

She states that the AI context changes many things and says there must be the political will to invest in indicators that measure where women are positioned in the structural transformation of economies and societies driven by AI . She concludes that this is not just about connecting women to the ICT market, but about ensuring meaningful pathways to empowerment .

The speaker argues that mapping AI capacity-building initiatives is useful, but it does not tell us whether AI systems are safe, contestable or open to redress when harms occur. The core point is that accountability metrics need to assess risks and remedies, not just the existence of training or capacity programmes.

She refers to the WSIS Plus 20 outcome assigning the UN Interagency Working Group on AI the task of mapping existing capacity-building initiatives and reporting to the global dialogue on AI governance . She then argues that while this is useful groundwork, knowing which countries have AI training programmes does not show whether deployed systems are safe, contestable or subject to redress, and she suggests the indicator review should include a specific accountability track that measures harms and redress mechanisms .

The speaker argues that cyber misuse is systematically under-measured because malicious actors deliberately hide their activities. Therefore, measurement systems that depend mainly on state self-reporting are structurally incapable of capturing many of the most serious digital harms.

She says the misuse of digital technologies has been structurally absent from measurement and sharpens this by saying it is 'invisible by design' because both state and non-state malicious cyber actors have incentives to remain undisclosed . She concludes that indicators based only on state self-reporting will always undercount the harms that matter most .

The speaker argues that cyber accountability should focus on the human consequences of attacks rather than only technical incident counts. In her view, counting breaches without measuring their effects on civilians and institutions misses the real social impact of cyber harm.

Drawing on previous work at the Cyber Peace Institute and Protect NGO, she says they built public platforms showing how cyber attacks affect civilian infrastructure such as hospitals and local governments in peace and conflict, as well as civil society and humanitarian organisations . She adds that this work showed that merely counting incidents is insufficient because a breach is not just a technical event, and that methodologies should instead measure the human harm behind the attacks, something current WSIS indicators do not capture .

The speaker argues that WSIS does not need to design AI and cyber accountability metrics from scratch. Instead, it should formalise and integrate existing civil society and expert tools into the WSIS review and reporting process.

She proposes public dashboards that track commitments against actual outcomes in real time, combined with independent evaluation by civil society or third-party experts and clear reporting obligations for all actors, not only states . She also points to examples such as the Internet Accountability Compass and says AI and cyber accountability metrics could similarly be built by integrating one or two existing tools directly into the review as practical test cases for more formal WSIS reporting .

Desiree Miloshevic Evans argues that the technical community can contribute practical, open and real-time data that improve WSIS monitoring, particularly on network resilience and shutdowns. She suggests that these tools can complement more traditional indicators by showing what is actually happening on networks during disruptions.

She says RIPE NCC and ICANN can organise Internet measurement days and that RIPE NCC runs open data tools such as RIPE Atlas, with around 12,000 probes that anyone can install on their network . She explains that these tools help measure Internet quality, pathways and traffic, including during shutdowns or disruptions, and mentions additional services such as IP stat as well as the willingness to share these datasets with stakeholders and regulators interested in network resilience .