IGF 2023 DC-IoT Progressing Global Good Practice for the Internet of Things

Tuesday, 10th October, 2023 (23:45 UTC) - Wednesday, 11th October, 2023 (01:15 UTC)
WS 11 – Room J
Dynamic Coalition on the Internet of Things

Round Table - 90 Min


This session is organized by DC IoT in collaboration with DC CIV and with support of DC IS3C and open to collaborate with other IGF Policy Networks and Dynamic Coalitions, where appropriate. Our aim is to advance a clear global sense of Good Practice on IoT

This builds upon the “Internet of Things Good Practice Principle” that was recognized by the DC IoT during the IGF 2019:

Internet of Things Good Practice aims at developing loT systems, products, and services taking ethical considerations into account from the outset, in the development, deployment and use phases of the life cycle, thus finding an ethical, sustainable way ahead using loT to help to create a free, secure and rights enabling based environment: a future we want.

For our session at the IGF 2023 we plan for a roundtable and invited the following contributors to provide input to a moderated discussion:

  • Vint Cerf, Chief Internet Evangelist, Google – for his view of global good practice to optimize the use of the IoT and mitigate associated risks by promoting standards, interoperability and critical thinking;
  • Orly Lobel, Professor of Law at San Diego University – to focus on the impact of AI on the IoT, and how technology can make a difference for the benefit of society in an inclusive way (e.g., Tech4Good). She calls on policy makers to adopt a comparative perspective that recognizes human fallibility and the flaws and advantages of new digital technology.
  • Hiroshi Esaki, Professor at Tokyo University and President of JPNIC –addressing the role of IoT in buildings factories, campuses and cities (including power infrastructures), in the context of global infrastructure design data center and AI and cyber security for IoT. This is based on a session held during APrIGF.
  • Sarah T. Kiden, Design researcher at Northumberland University - on co-designing IoT with local users and supporting communities to ensure that the wider implications of IoT technologies are taken into appropriate account to shape and influence future deployments and how to embed values like trust, safety and integrity into the design processes of IoT devices. This is based on a session held during Africa IGF.
  • Alejandro Pisanty/Olivier Crépin-Leblond, DC CIV – on a Core Internet Values perspective in which concepts like best-effort, interoperability, openness, robustness, decentralization, end-to-end, scalability and permissionless innovation both affect and are affected by the expansion of the IoT in different scenarios (e.g., whether the IoT’s development is inertial, less regulated or more strongly regulated than present.) The increased prevalence of AI tools within the IoT is an emerging challenge. This was as reflected in the discussion during EuroDIG.
  • Narelle Clark, Internet Association of Australia – on labeling and certification as means to inform users on the use and safety of IoT components, devices, systems and services. This is based on a session held during APrIGF.
  • Lucien Castex, AFNIC  – on identity management and encryption as key factor in cybersecurity of LoRa and other IoT networks (in line with the zero-trust approach). This has been discussed during the Africa IGF session;
  • Wout de Natris, IS3C coordinator – on current policy and legislation practice around the world regarding IoT. This is based on a study in 2022, and discussion during EuroDIG.
  • Jonathan Cave, Warwick University and Alan Turing Institute – on data protection and privacy elements to be addressed in IoT environments. This is based on a session held during Africa IGF.

The Outcome of this IGF 2023 roundtable will be an issues document on lessons learned from 15 years of global engagement of the DC IoT within the IGF setting. Our aim is to use this as the basis for developing a publication to be published prior to the Summit of the Future in September 2024, with the – provisional – title: 15 Years of IOT Discussions within the IGF: Where we are and where we should go from here have to go". 

  • Maarten Botterman, ICANN Board; GNKS Consult; DC IoT Chair; Netherlands; 
  • Jonathan Cave, University of Warwick, DC IoT; UK;
  • Dan Caprio, Providence Group, DC IoT; USA;
  • Olivier Crépin-Leblond, DC CIV Chair; France;
  • Avri Doria, Technicalities, ICANN Board, DC IoT; USA;
  • Sarah T. Kiden, Northumbria University, ICANN ALAC; Uganda
  • Wolfgang Kleinwaechter, University of Aarhus, former DC IoT Chair ; Germany   
  • Wout de Natris, de Natris Consult, DC IS3C Chair; Netherlands
  • Shane Tews, Logan Circle Strategies, former DC IoT Chair
  • Vint Cerf, Google, USA
  • Hiroshi Esaki, University of Tokyo, President JPNIC Board of Directors, Japan
  • Sarah T. Kiden, Northumbria University, ICANN ALAC, Uganda
  • Alejandro Pisanty/Olivier Crépin-Leblond ,DC CIV
  • Dan Caprio, The Providence Group, DC IoT, USA
  • Sandoche Balakrichenan

    , AFNIC, France

  • Wout de Natris, DC IS3C, The Netherlands 
  • Jonathan Cave, Warwick University and Alan Turing Institute, UK
Onsite Moderator
Maarten Botterman & Olivier Crépin-Leblond
Online Moderator
Avri Doria
Jonathan Cave (TBC)

2. Zero Hunger, 2.3, 3. Good Health and Well-Being, 4. Quality Education, 8. Decent Work and Economic Growth, 9. Industry, Innovation and Infrastructure, 11. Sustainable Cities and Communities, 12. Responsible Production and Consumption, 13. Climate Action, 14. Life Below Water, 15. Life on Land


-    Goal 2 - End hunger: specifically 2.3 double the agricultural productivity and incomes of small-scale food producers) but also 2.1 and 2.2, in particular by providing access to information via the Internet in local languages, and providing opportunities to trade crop locally, and regionally, using Internet based trade platforms;

-    Goal 3 - Ensure healthy lives: by facilitating access to health knowledge, health warnings and care via the Internet in local languages, and local language scripts;

-    Goal 4 - Quality education: providing access to education and child care, and knowledge, via the Internet in local languages and local language scripts;

-    Goal 8 - Economic growth and jobs: Sustainable economic growth will require societies to create the conditions that allow people to have quality jobs that stimulate the economy while not harming the environment. This will include access to training and knowledge for work, as well as access to the market place, and being able to do so via the Internet using local language and local language scripts will be key;

-    Goal 9 - Resilient infrastructure and sustainable industry: in particular 9.B Support domestic technology development, research and innovation in developing countries, and 9.C Significantly increase access to information and communications technology and strive to provide universal and affordable access to the Internet in least developed countries by 2020, as local language use and local language script use via the Internet will be a key enabler;

-    Goal 11 - Sustainable cities and settlements: in particular 11.1 By 2030, ensure access for all to adequate, safe and affordable housing and basic services and upgrade slums – having access to services means people will need to be able to access it in their local language and local language scripts; and 11.3 By 2030, enhance inclusive and sustainable urbanization and capacity for participatory, integrated and sustainable human settlement planning and management in all countries – for participation to be possible for  all, local languages and local language scripts are a sine qua non;

-    Goal 12 - Sustainable consumption and production: in particular 12.8 By 2030, ensure that people everywhere have the relevant information and awareness for sustainable development and lifestyles in harmony with nature will require access to such information via the Internet in local languages and local language scripts. Same is also true for Goal 13 Climate action; and 16 Peace, justice and strong institutions with regards to broad access to information.

-    Goal 13, 14 and 15 Sustainable environment related: measuring and feedback loops;

Key Takeaways (* deadline 2 hours after session)
When using IoT devices and services, strong identification becomes key to protect these from tampering. This identification may be between devices, for instance those that together provide a service, or form together a so-called “cyber physical system” such as a car, a house, an airplane, etc. When this identification is between people and devices, there needs to be sufficient measures in place to ensure privacy by default.

With the ongoing growth of IoT deployment throughout our world, scaling issues are important to consider. Going forward to design imperatives need to be taken on board: (1) security by design - every device needs to be protectable (and updatable when needed); and (2) every device needs to be as carbon neutral as possible (as there will be many, including those that are dependent on power).

Call to Action (* deadline 2 hours after session)

Require appropriate security measures for IoT devices that can be handled by those that use them, and ensure appropriate labeling (dynamic for those devices that are software updatable) to make it possible for user to assess the risks and take the necessary measures.

Set global standards for this, as it concerns devices that are developed all over the world, and are deployed all over the world. National/regional initiatives will need to take global good practice into account.

Session Report (* deadline 26 October) - click on the ? symbol for instructions

IGF 2023 DC-IoT Progressing Global Good Practice for the Internet of Things

The session considered IoT governance from various perspectives. To understand baseline IoT evolution, associated challenges, opportunities and responses, the IoT could best be understood as an internet of data, devices, systems or functions. For simplicity, we can call these “Internets of X” (IoX). Each perspective brings its understanding of what is possible, desirable or undesirable and tools and processes needed for governance.

Each approach must be considered in its own terms, but they start from a common base of experience and must ultimately come together to provide good governance. This leads to the need for an ecosystem comprising of stakeholders such as technical experts, governments, service providers, manufacturers, users, standards bodies, military vs civilian organisations, etc., varying in global and regional perspectives.

One immediate consequence is that IoT governance must respect a range of perspectives. Our fundamental principles are unlikely to be universal, especially when applied to specific IoT contexts. By analogy with the sensors and actuators of the IoT itself, governance needs to ‘sense’ the interests and perspectives of all significantly affected parties and somehow balance them to inform decisions at various levels. In other words, it requires multistakeholderism. It is not that specific expert groups (e.g., engineers) are insensitive to the needs of others (e.g., end users) but that they may misunderstand their interests, capabilities and behaviour.

The session began with a consideration of simple and recognisable use cases in which major challenges can already be seen (though they will become more complex). IoX components and their complex or hybrid assemblages will and should interact with others, so they must be identified uniquely and discovered with appropriate levels of precision, reliability, and permanence and be capable of enrolment in or separation from IoX systems. The concept of ‘identity’ has some subtlety. For instance, a smart home must be able to recognise and be recognised by new IoT components added to the system on a permanent or temporary basis, accorded the right kinds of access and privileges and tracked or remembered appropriately. These identities enable necessary functions, including the granting of trust. But they need not be unique, durable or universal. Indeed, categorical or shared identities (e.g., type certification) may be more practicable, scalable, flexible, future-proof, secure and robust to, e.g., (hardware, software or data) updates and interconnection or federation to create identifiable hybrid systems. Three subtleties linked to identity that came up in the discussion were security (including but not limited to cybersecurity), privacy (including but not limited to data privacy) and ownership (including protections against identity theft or misuse and, conversely, the use of identity to carry liability or responsibility).

Various identity schemes were discussed, ranging from central registries of semi-permanent discrete identities (along the lines of the DNS model) to purely transactional or temporary mutual authentication and identification schemes. These have advantages and drawbacks ranging from theoretical to practical, including technical, legal, commercial, security and other considerations. No single approach seemed to fit all foreseeable circumstances. In placing these in context, the panel recognised that the same concepts applied to the human beings (and organisations) that create, operate and use the IoX. For example, a person is more important than devices or data attributed to him/her, and human rights and responsibilities (e.g., of association and expression) cannot safely be extended to, say, their smart digital assistants. This cuts two ways; it may not be useful to hold a human being accountable for what their devices do in response to interactions with other systems, which the ‘user’ may not even perceive, let alone understand or control. Conversely, the automation of routine functions may result in their receiving less considered and responsible human attention, with unintended, undesirable and possibly irreversible results.

The discussion also considered desirable properties that might provide an ethical framework for IoT governance. Many are familiar, e.g., interoperability, transparency and accountability, robustness, resilience, trustworthiness, user empowerment, privacy and security. They are not IoT-specific but may need to be reinterpreted in that context. For example, IoT devices can harvest a wide range of data almost invisibly, which creates general privacy and security risks and affects global development, e.g., via ‘data colonialism’ whereby devices originating in and provisioned by the global north can be used to capture data from users in the global south to produce innovations for the benefit of the north and to lock in users in the south in ways that inhibit their techno-societal development.

One desideratum came up in relation to technologies, service provision, use cases, data issues, labelling and certification schemes and legal frameworks, and scalability. This is a generic issue, but the panel highlighted aspects that stand out clearly in the IoT context. One is complexity; as systems scale quantitatively, their qualitative properties may change and, with them, the appropriate kind of governance. Rules may need to be more general, neutral, principles- or function-based. Alternatively, governance may need to move between the data, device, software, etc., planes as systems interconnect in larger and more diverse ways. Another is practicability; effective governance may require limits on scale or interoperability. A further aspect is Quality of Service (QoS). The IoT-specific emphasis on low latency can constrain system scale, security or flexibility. Beyond this, QoS considerations may lead to multi-tier systems, which may reduce economic welfare, hinder interoperability or distort innovation. Large-scale systems may also be more susceptible to intentional or accidental compromise; effective access control in large environments may lead to inappropriate inclusions or exclusions. Under laissez-faire evolution, IoT systems may reach stable sizes and configurations, but these may not be optimal. Finally, very large systems may be difficult to govern with national or self-regulatory arrangements. For example, identification and certification schemes that identify individual devices or types scale with their number but cannot identify even pairwise interactions (which scale as the square of the number of interacting entities). As scale increases, management overloads, costs increase, and utility and use eventually decline. This, however, depends on the governance architecture; a centralised system (analogous to the cloud) offers economies of scale (or diseconomies) and a natural platform for observing systemic behaviour and emergent threats (if not weak signals). However, it creates additional power asymmetries and vulnerabilities; no one governance architecture will likely fit all cases. The group also mentioned other aspects of scale, such as environmental impact.

Another aspect that ran through the various phases of the discussion was trust and trustworthiness; beyond the customary discussion of e-trust, the panel contrasted high-trust and Zero-trust approaches to the problems of identification and interoperability.

The issue of AI in the IoT comes up often but not in depth. The panel recognised that it complicated the IoT, especially when considering smart devices and the emergent intelligence of connected systems. Foreseeability and explicability were discussed, as was the possibility that data-driven systems might be particularly vulnerable to noisy or biased data.

The panel considered various legal approaches and the ‘regulatory game’ being played out among countries, industries and civil society groups. Governance competition could spur the development of innovative and effective standards if different approaches can be compared and a suitable global standard emerges through a kind of ‘Brussels Effect’. This seems more promising than a too-rapid imposition of global standards and regulations whose implications cannot be foreseen. However, this result is not guaranteed; we could see damaging fragmentation or a rich diversity of approaches matching different contexts. Research on policy initiatives in 40 countries around the world shows that governments often do not regard modern global open source standards and global good practices with security at the core as “important”. It was suggested that governments could lead the way by taking such standards actively on board in their procurement activities. Keeping the discussion going and actively engaging with other DCs guarantees a positive outcome and an increased understanding of good global practices in IoT governance. Three important takeaways:


IoT data, especially AI-enhanced, should be understandable, accessible, interoperable, reusable, up-to-date and clear regarding provenance, quality and potential bias.


At the level of devices, there need to be robust mechanisms for finding, labelling, authenticating and trusting devices (and classes of devices). These should survive retraining, replacement or updating but be removable when necessary for functional, security or privacy reasons. To ensure IoT functionality, trustworthiness and resilience, market information and incentives should be aligned. Labels provide a powerful tool; many countries have developed and adopted IoT trust marks, and the time has come to start working towards their international harmonisation.


Functions are not all confined to single devices, designed in or provided by system integrators; they can also be discovered by end-users or emerge from complex system interactions in cyber-physical systems (CPS) and IoT-enabled services. Governance requires methods for recognising, protecting and controlling these functions and their impacts.