Speaker 1: Nozha Boujemaa, Technical Community, Western European and Others Group (WEOG)
Speaker 2: Hodes Cyrus, Government, Asia-Pacific Group
Speaker 3: Huan Luke, Private Sector, Asia-Pacific Group
Speaker 4: Olivier Clatz, Private Sector, Western European and Others Group (WEOG)
Speaker 5: Anne Carblanc, Intergovernmental Organization, Western European and Others Group (WEOG)
Panel - 90 Min
A major value-add of this panel is the diversity of perspectives among panelists, which span national and international governments, international organization, non-profit, and industry. The panel format allows each panelist sufficient time to express their views on the workshop topic’ as well as the opportunity to debate and discuss this from different perspectives. In this respect, the panel format is optimal to elicit their diverse views and expertise.
Initially each panelist will have 5 minutes to make an opening statement, including the need, challenges, and concrete strategies in for “Maturing efficient processes for global governance of AI.’ Afterwards, the panel moderator will coordinate discussion and debate among panelists to ensure equitable speaking time. The moderator will also balance panelists’ questions and comments with those of audience members in the 50 minute discussion period. The panel will finalize with concluding (1-2 minutes) remarks from each of the panelists summarizing the open areas of discussion and concrete recommendations.
The workshop will comprise of speakers from a wide range of backgrounds. We are committed to bringing diversity in the discussion of AI governance, given that the topic has widespread implications for all of humanity. As such, we have intentionally assembled a speaker line up which is gender balanced (2 female/ 3 male), spans the globe with speakers from China, UAE and Western Europe, representative of the multi-stakeholders from civil society, academia, government and industry that are important in shaping AI governance. We have also made concerted effort to include first time speakers to this IGF session so as to expand the network reach and garner broad perspectives during the workshop.
The panel will address the question of ‘How can we mature efficient global governance processes for AI?’ During the workshop, the moderator will encourage panelists and audience members to discuss the rise, dynamics, needs, challenges, and concrete strategies for achieving global governance of artificial intelligence.
First, the moderator briefly introduces each of the 5 panelists after which they each provide 5-minute opening remarks (30 minutes).
Next, the moderator coordinates discussion & debate among panelists and audience members (50 minutes).
Lastly, panelists provide 1-2 minute closing remarks summarizing their perspectives and offering concrete strategies and a path towards achieving global governance in artificial intelligence (10 minutes).
The moderator aims to balance breadth with depth of topics while also achieving engaging discussion and debate. Topics include:
What is “global governance” in the context of AI?
Why is global governance important to govern the rise of AI?
Who are the key stakeholders and actors needed in global governance, and what are their respective roles?
What are concrete strategies & platforms to achieve multi-stakeholder governance (e.g. government, industry, non-profits)?
What are current challenges to global governance of AI?
What are concrete strategies, methods, platforms or institutions to coordinate global governance of AI?
The topics above are purposefully general to elicit broad perspectives and insights. However, we expect participants to contribute specific and innovative policy recommendations for the global governance of AI, such as new platforms for international cooperation, leveraging other emerging technologies to achieve governance (e.g. distributed ledgers or blockchains), and leading theories in governance (e.g. realism and balancing ‘soft law’ with ‘hard law’). The panelists’ diverse sectoral positions will contribute valuable insight into the differing objectives and constraints faced by governments, industry, and non-profits. Panelists will explore new models of cooperation to leverage stakeholders’ varying capacities and competencies and to overcome their unique shortcomings and constraints.
A key goal is facilitating lively discussion and debate. The objective is to achieve both engaging discussion and diverse perspectives, from the diverse backgrounds of the panelists and audience members. For example, panelists’ diverse professional backgrounds will naturally place them on different sides of some discussions.
The majority of the session (50 out of 90 minutes) will be dedicated to facilitating debate and discussion raised by the moderator, audience members, and remote audience members. The moderator will be prepared with a set of initial questions designed to generate debate and discussion among panelists. During panel discussion, the moderator will call on all participants to put forward questions and comments to the panelists. This format allows showcasing of panelists’ diverse and expert opinions while also encouraging discussion with audience members. After a round of discussion, the moderator will take new questions from audience members including remote audience members. To facilitate this, the remote moderator will gather questions submitted online and send them via an online platform to the moderator.
Policy Question: How can we mature efficient global governance processes for AI?
Artificial intelligence (AI) is poised to transform the economy, society, and political systems we know today. The AI revolution’s impact inextricably combines substantive opportunities and serious societal risks. Beneficial outcomes include a wave of productivity gains, hyper-tailored education, new drug discovery, safer roads, and efficient energy usage, among others. On the other hand, AI brings the spectre of widening inequality and mass unemployment, the loss of human privacy, dignity, fairness and agency. Governance frameworks and processes that are global and collectively built are critical in shaping the trajectory of the AI revolution which maximizes the upsides and minimizes the downside risks for everyone.
The policy question that this workshop will explore is “How can we mature efficient global governance processes for AI?” This topic has been continually worked on by The Future Society, with recent examples such as the Global Civic Debate on “Governing the Rise of AI”. This was a trans-civilisational collective intelligence initiative that sought to confront and align perspectives from US, China, Europe, Japan and other countries, on the rise, dynamics and consequences of the AI revolution. This initiative, conducted through a curated online discussion forum (www.aicivicdebate.org) over 7 months, resulted in deep insights gathered from over 2,200 participants that made more than 3,300 proposals for actionable policy recommendations to governments across the world.
Building on this, The Future Society also led the organisation of the first roundtable on the “Global Governance of AI” at the 2018 World Government Summit, commissioned by the UAE Minister of State for AI. We convened over 100 global AI experts, scientists, practitioners, investors and regulators, from over 20 countries gathered in the UAE for the day-long forum to facilitate dynamic policy recommendations and create a roadmap for ultimate adoption by nations around the world. The result of this roundtable was a highly visible and actionable report and action plan delivered to the UAE Minister of State for AI, to create a revolving global platform for AI governance. This forthcoming proposed workshop to the Internet Governance Forum naturally advances the multitude of efforts we have been making to shape rigorous and inclusive discussions on AI governance.
Several competitive dynamics are unique to AI and necessitate global governance and coordination. Competition among global firms and nation-states in a ‘global race’ to develop AI may accelerates innovation at the expense of ethical and safety standards. Differences in regulatory and consumer protection regimes lead to potentially destructive imbalances between countries with higher ‘risk appetite’ for growth and innovation in order to enable greater access to goods and services, such as the case in India, and those aiming to protect citizens from potential threats to privacy and security, as in Europe. Such global frictions between different value systems and socio-economic fundamentals require proper governance and management. Without mature coordinated processes of global governance, the AI revolution could manifest in a race to the ethical bottom rather than create incentives for upholding social and ethical standards among national and private players.
Governance of AI will also benefit from a complementary mix of ‘soft law’ (e.g. industry guidelines, codes of conduct and practices, technical standards) and ‘hard law’ (e.g. regulation). ‘Soft law’ and ‘hard law’ are implemented in varying time scales, engage different communities of practice and interest, and thus result in differing policy impacts. Therefore, to be both inclusive and effective, the governance of AI should incorporate elements of multi-stakeholder and interdisciplinary, reaching across government, industry, non-profits, academia, and civil society. The proposed panelists for our workshop offer such diverse perspectives, providing the opportunity to make ground towards holistic and representative global governance of AI.
The current AI revolution can be defined as the convergence of three trends; big data, machine learning algorithms and cloud computing power, which are nested within a complex socio-technical system. As such, the governance of AI is underpinned and inextricably connected to the collection, storage, processing and consumption of data. In order to better understand the dynamics and mechanisms of AI governance, the panel will also share perspectives on data governance where data represent the real asset but could be a source of vulnerability for the whole system. New solutions will also be explored that ensure the use of data maintains individual privacy, security, representation and access, such as ‘Data Commons’, ‘AI Commons’ or Blockchain to support securitization.
Finally, the panel will observe how governance in the AI revolution may reach beyond ‘governance of algorithms’ to include ‘governance through algorithms’ and the fact that AI is part of the solution. The scale, velocity, complexity and the systemic diversity of AI mandates new forms of governance (audit, policing, investigation…) whereby we use AI to govern ourselves, as much as we govern AI.
While the trajectory of AI development is uncertain, its potential for step-change impact on humanity is clear. There is ample case for the application and deployment of AI technologies to monitor, serve and overcome society’s greatest challenges including all of the the United Nations Sustainable Development Goals. The path and impact of the AI revolution will largely depend on governance and policy. Unique dynamics in the rise of AI, including fervent competition to develop AI as rapidly as possible and imbalances in regulation and ‘risk appetites’ may amplify downside effects. This panel charters the course for a collaborative discussion among global stakeholders to create a coordinate effort for managing the AI revolution in the most efficient and beneficial manner.
Organizers will arrange a remote/online platform for remote audience members to listen to the panel discussion and engage by submitting questions to the panelists. A designated remote moderator from the organizers’ team (The Future Society and DATAIA) will gather questions submitted online and transfer them via an online platform to the moderator during the panel discussions. The panel moderator will include several remote audience questions during the 50-minute discussion period.