The Expansion of AI Governance

The World Economic Forum has recently announced its AI Governance Alliance . As the World Economic Forum describes it:

A pioneering multi-stakeholder initiative that unites industry leaders, governments, academic institutions, and civil society organizations, to champion responsible global design and release of transparent and inclusive AI systems.

Initiatives like this are critical to the ethical implementation of AI in society. I spoke about this at the 2023 Berlin Global Dialogue along with my colleagues Madalina Bouros and Margarita Fadeeva. In our presentation, we highlighted the importance of bringing together a diverse set of representatives across society in order to represent the various perspectives of those who may be impacted by AI. Efforts like the AI Global Alliance are critical for the safe and ethical expansion of AI.

I’ve included our remarks below for those who may be interested.?

_____________________________________________________________________

Ethical and Manageable AI for Good Versus the End of Human-Dominated History: What's Our Choice?

While not a novel technology, AI has witnessed unprecedented democratisation and acceleration in development thanks to advancements like generative AI, epitomised by technologies like Chat GPT. This burgeoning landscape presents us not only with opportunities but also with ethical dilemmas and challenges that demand our attention and resolution.

As a global community, we need to be aware of both the opportunities AI provides. But also, we need to be prepared for the problems such a large change in society is bound to bring. We must find a balance where we both promote AI use but also ensure the decisions made around AI are responsible, ethical, and aligned with our shared values.

AI's democratisation empowers the average citizen with robust capabilities while propelling the international marketplace's AI proficiency. However, with great power comes great responsibility. While AI acts as a transformative agent in society, it brings along potential pitfalls that require careful navigation. Challenges such as algorithm bias, algorithm hallucinations, ethical considerations in AI training, ambiguous effects on global mental health, and shifts in employment types due to AI-induced automation are prominent.

That is why we’d like to introduce our concept – The AURA (The Alliance for Universal Responsible AI) – a multi-stakeholder organisation promoting “AI for Good”.

As professionals, and also a Global MBA study group at the ESMT Berlin (European School of Management and Technology, a leading business school in Germany) we have been thinking about the implications of AI in the global economy and the current challenges.

Creating an International AI Organisation - AURA (Alliance for Universal Responsible AI) that serves as a platform for collaboration between government, business, academia and others can have transformative effects, particularly when the focus is on "AI for Good” for the generations to come.

The establishment of an AURA with a focus on AI for Good can drive global efforts to leverage technology for societal and environmental betterment. A multi-stakeholder approach, with active participation from government, business, and science communities, will be crucial to harness the full potential of AI in addressing the world's pressing challenges.

By having a well-structured operating model, the AURA (Alliance for Universal Responsible AI) can effectively work towards its mission of harnessing AI for societal and environmental betterment. This multi-faceted approach ensures that all stakeholders have a voice and that the organisation can adapt and evolve in a rapidly changing world.

By involving a broad and diverse range of stakeholders, AURA can ensure its efforts to promote responsible AI are comprehensive, balanced, and responsive to the needs and concerns of society at large. Let’s take a look at the AURA stakeholder groups.

Let me bring some concrete examples of how it could be structured.

The AURA (Alliance for Universal Responsible AI) Organisational structure (stakeholder groups):

Governments and Public Bodies

National Governments - Representatives from technology, education, labour, and finance ministries.

International Bodies: Representatives from the UN, World Bank, WTO, and regional bodies such as the EU, ASEAN, etc.

Regulatory Agencies - Institutions responsible for technology and data regulations, labour laws, and other related areas.

Technology and AI Industry

Tech Giants - Companies like Google, Facebook, Microsoft, and others are at the forefront of AI research and development.

Start-ups and Innovators - Entities that are developing cutting-edge AI solutions.

AI Researchers and Scientists -? Individuals from academia and industry with expertise in AI.

Civil Society and NGOs

AI Ethics Advocates - Groups focusing on ethical considerations of AI.

Labor Rights Groups - Advocates for workers' rights in the face of AI-induced job transformations.

Consumer Protection Agencies - Entities ensuring AI technologies are used responsibly with consumer interests in mind.

Educational and Research Institutions

Universities - Particularly those with strong technology, social science, and humanities programs.

Research Bodies - Institutions focusing on the study of AI, its implications, and its potential for societal benefit.

Business and Industry Groups

Industry Associations - Such as chambers of commerce or industry-specific groups.

Labour Unions - Representing workers across various sectors.

Innovation Hubs - Tech incubators, accelerators, and other entities fostering AI innovation.

Investment and Financial Sector

Venture Capitalists - Individuals or firms investing in AI start-ups.

Philanthropists and Impact Investors - Those keen on driving positive societal impact through AI.

Media and Communication

Tech Journalists - Individuals who communicate the latest in AI to the public.

Public Relations Professionals - Those who can help shape the narrative around AURA's mission and objectives.

The General Public and Communities

Community Leaders -? Those who can provide insights into the needs and concerns of local communities.

General Citizens - Providing a broader perspective and grounding the initiative in real-world concerns.

Independent Auditors and Accountability Bodies

Third-party Audit Firms - Ensuring transparency and responsibility in AI endeavours.

Certification Bodies - Offering certifications for AI projects adhering to AURA's guidelines and standards.

Global Think Tanks and Experts

Futurists - Experts who forecast trends in AI and its potential implications.

Historians & Sociologists - Offering insights into historical precedents and social dynamics.

Let’s summarise the Operation Model components.

Leadership Council - This could comprise leaders from government, science, and business communities. This council oversees the strategic direction of the organisation.

Operational Teams - Specific teams dedicated to each of the concepts (like Repository Management, Partnership Coordination, Think Tank Operations, etc.)

Advisory Boards - Specialised groups that offer insights into ethics, technology, public policy, and more.

Stakeholder engagement key activities

Regular meetings, workshops, and consultations with government officials, business leaders, scientists, and civil society. Online platforms for continuous collaboration, knowledge sharing, and joint decision-making.

Partnerships & Collaborations

We would recommend establishing Memorandums of Understanding (MoUs) with key organisations, universities, and businesses worldwide. Joint ventures on specific AI projects with organisations that share the IAIO's objectives.

Communication Strategy

Regular publications, press releases, and updates are important to keep the global community informed. An active presence on social media, webinars, and global conferences to spread the mission and achievements of the organisation. Periodic publications of research, case studies, and white papers will help establish effective communication.

The AURA (Alliance for Universal Responsible AI) Governance model:

Let’s take a look at the AURA Governance model.

Each governance body within AURA is equipped with its own terms of reference, operational guidelines, and frameworks for decision-making. The inclusive engagement of these committees and teams allows AURA to operate effectively and transparently, championing responsible AI that serves the greater good.

Executive Committee (EC)

Potential role - Provides strategic direction and ensures that AURA's vision and mission are realized.

Governance composition - Representatives from each stakeholder group to ensure comprehensive perspectives.

Key decisions and responsibilities: Budget allocation, policy formulation, and high-level strategic decisions.?

Operational Management Team (OMT)

Potential role - Oversees the day-to-day operations and implementation of AURA's projects and initiatives.

Governance composition - Operational leads for each AURA's structured function (e.g., Unified AI Good Repository, Ethical AI Frameworks).

Key decisions and responsibilities - Project approvals, timeline decisions, and resource allocations.

Stakeholder Advisory Board (SAB)

Potential role - Provide insights, advice, and expertise to the Executive Committee based on each stakeholder's perspective.

Governance composition - High-level representatives from each of the ten stakeholder groups.

Key decisions & Responsibilities - Guiding the priorities of AURA based on feedback from each stakeholder group.

Ethics and Compliance Committee (ECC)

Potential role -Ensures all AI-related initiatives adhere to ethical, transparent, and compliance standards.

Governance composition - AI Ethics Advocates, Certification Bodies, and members of Regulatory Agencies.

Key decisions & Responsibilities - Certification and approval of AI projects, development of ethical guidelines.?

Education and Outreach Team (EOT)

Potential role - Handles all systemic education, enlightenment work, and debunking of myths and rumours about AI.

Governance composition - Educators, tech journalists, PR professionals, and representatives from universities.

Key decisions and responsibilities - Curriculum and content development, workshop scheduling, and outreach programs.

Funding and Grants Committee (FGC)

Potential role - Oversees funding opportunities and grant distribution.

Governance composition - Venture capitalists, philanthropists, and representatives from public and private financial bodies.

Key decisions and responsibilities - Allocation of funds, grant approvals, and funding partnerships.

Audit and Review Team (ART)

Potential role - Ensures transparency in AI algorithms and conducts regular audits.

Governance composition - Third-party audit firms, AI experts, and certification bodies.

Key decisions and responsibilities - Audit schedules, transparency report releases, and algorithmic compliance checks.

Policy and Regulatory Affairs Team (PRAT)

Potential role - Shapes policies, and regulations, and engages with policymakers on implications, challenges, and opportunities of AI.

Governance composition - Policymakers, tech leaders, business leaders, and representatives from civil society.

Key decisions & Responsibilities - Policy recommendations, regulatory adjustments, and public policy dialogues.

Research and Innovation Council (RIC)

Potential role - Guides AI research, identifies global challenges and fosters innovation.

Governance composition - AI researchers, scientists, futurists, and representatives from innovation hubs.

Key decisions & Responsibilities - Research priorities, challenge identification, and innovation partnerships.

Awards and Recognition Committee (ARC)

Potential role - Recognizes and rewards significant contributions to the AI for Good movement.

Governance composition - Notable figures in the AI world, historians, sociologists, and other relevant stakeholders.

Key decisions and responsibilities - Award categories, selection criteria, and recognition events.?

The AURA (Alliance for Universal Responsible AI) Organization & Execution (Operational Mechanisms):

The AURA (Alliance for Universal Responsible AI) Organization and its meticulously designed operational mechanisms, serve as the bedrock for its execution and functionality. The formation of AURA represents a collaborative endeavour aimed at nurturing responsible and universally beneficial AI technologies. This alliance not only underscores the commitment to crafting advanced AI solutions but also to doing so in a way that is ethical, transparent, and inclusive. The ensuing sections will unveil the intricacies of AURA’s organizational structure and the operational mechanisms it employs to foster an environment where responsible AI can thrive and contribute positively to the global community.

Tech and Policy Dialogue Forums

We would also recommend organized forums where policymakers, technologists, business leaders, and civil society can discuss the implications, challenges, and opportunities of AI for Good, shaped policies and regulations that nurture and promote AI for Good initiatives.

Public-Private Partnership Initiatives

Facilitated collaborations between governments, private businesses, and research institutions to co-fund and co-develop AI for Good projects will support effective co-partnership. The creation of mechanisms for pooling resources, expertise, and data will facilitate growth.

Global Challenges Think Tank

This section can identify global challenges (e.g., climate change, health crises, food security) where AI can make a significant difference and assemble multi-disciplinary teams to brainstorm, develop, and deploy AI solutions tailored to these challenges.

Ethical AI Frameworks and Standards

The highest priority would be the formulation of guidelines that ensure AI for Good projects prioritize ethics, fairness, transparency, and inclusivity, but also establishing benchmarks and best practices for projects to align with human rights and environmental standards.

Innovation Hubs and Labs

Another importance will be to establish physical or virtual spaces where researchers, developers, and innovators can collaborate on AI for Good projects. These hubs can offer resources, mentorship, and the necessary infrastructure for rapid prototyping and testing.

Education and Capacity Building

This model could offer training programs tailored to the integration of AI in various sectors (e.g., healthcare, agriculture, education). Hosting workshops, webinars, and conferences highlighting the potential of AI for Good and sharing success stories.

Inclusivity and Representation

The governance will ensure representation from diverse regions, cultures, genders, and socio-economic backgrounds in decision-making, project development, and implementation. Promote projects that specifically cater to marginalized or underserved communities.

Global AI for Good Awards

The steering committee will recognize and reward individuals, teams, or organizations that have made significant contributions to the AI for Good movement. This can motivate others and offer successful initiatives a platform to scale.

Proposed Operational Mechanisms key activities

Annual Planning - Setting yearly goals, budgets, and focus areas based on global trends and needs.

Quarterly Reviews - Assessing progress against goals, understanding roadblocks, and readjusting strategies.

Feedback Loops - Mechanisms to collect feedback from stakeholders and the global community, ensuring the organization remains adaptive and responsive.

The AURA (Alliance for Universal Responsible AI) Technology Infrastructure: However, it's not just about algorithms and codes. It's about understanding, mutual control, and security systems. Here are the main components.

Cloud-based Repository - For the Unified AI Good Repository, ensuring accessibility and scalability.

Collaboration Tools - Platforms for video conferencing, project management, and real-time collaboration.

Analytics Platforms - Tools to measure the impact, reach, and success of AI for Good projects.

Unified AI for Good Repository - A centralized database of AI research, projects, tools, and datasets dedicated to solving global challenges, ensuring easy access and discoverability. Integration of case studies and real-world applications that have had positive societal impacts.

The AURA (Alliance for Universal Responsible AI) Global Radar:

Impact Measurement and Analytics:

The goal is to develop standardized metrics to evaluate the societal, environmental, and economic impact of AI for Good projects. The purpose is to provide tools and platforms that projects can use to assess and report their impact.

Grants and Funding Opportunities:

The goal is to establish a grant system to fund promising AI for Good projects, especially those in underfunded regions or sectors. The purpose is to facilitate partnerships between philanthropists, investors, and project developers to ensure sustainable funding.

Crisis Management & Response:

The goal is to dedicated team and protocols for managing crises, ensuring quick and effective responses. Regular drills and scenario planning to be prepared for any unforeseen events.

Risk Management:

The goal is to identify potential risks, from technology failures to geopolitical challenges. The purpose is to have mitigation strategies in place, ensuring the organization can navigate any challenges it encounters.

Certification of AI professionals:

Global AI Professional Certification (GAIPC)

In today's rapidly evolving professional landscape, where career paths are non-linear, and professionals frequently switch roles and industries, there's a growing need for an internationally recognized certification program for AI professionals. The "Global AI Professional Certification (GAIPC)" program aims to address this need by offering a comprehensive and adaptable certification framework for individuals in the field of artificial intelligence.?

AURA is a global non-profit organisational platform (concept) that focuses on bringing stakeholders in AI development together. AURA will enable collaboration - between governments, businesses, and academia… among others - to leverage AI for social good.

Our Goal at AURA is to bring key stakeholders in AI together through a platform-based approach, and deliberate collaboration to harness the full potential of AI in addressing the world's challenges.

AI implementation itself presents challenges and we must work together to mitigate risk in AI implementation as we promote AI adoption globally. Working collaboratively towards 'AI for Good'."

We, as a global community, need to come together to collaborate, educate, and innovate, and ensure that the AI era… is marked by global responsibility for the benefit of humanity. Concepts like AURA are a critical component in creating that future.?

The concept was introduced for the first time at the Berlin Global Dialog in September 2023 during the panel discussion (Responsible AI: From Theory to Action to Solve Global Challenges). Berlin Global Dialogue is a new international forum that unites leaders from business and policy to find joint solutions for a global economy in transition.