Overview

Aligning artificial
intelligence

By AI systems, we refer to the OECD definition, concerning not only narrow AI systems designed for discrete tasks, but also general-purpose AI systems with more versatile capabilities.

through better
governance

By governance, we mean the ‘rules of the game’—the formal and informal mechanisms, including laws, policies, and norms, that affect the behaviors of actors in the AI value chain.

Our impact

>8,000

senior decision-makers engaged

>30,000

individuals involved

>4,000

students taught

>60

institutional partners

>120

countries reached

Across our public events, we have engaged over 8,000 senior decision-makers, including legislators, industry representatives, academics, civil society advocates, and regulators, as well as over 30,000 individuals from around the world. Through our educational programs, we have taught 4,000 students—including judicial operators and senior executives from the public and private sectors—about various aspects of AI governance. This would not have been possible without over 60 institutional partners, including OECD, UNESCO, Harvard Kennedy School, the Patrick J. McGovern Foundation, the Global Partnership on AI, and the Future of Life Institute. As the governance of AI concerns us all, we recognize that global diversity is crucial. This is reflected not only in our team but also in our engagement with individuals from over 120 countries.

Our guiding tenets

Practicality

We prioritize ideas by their potential to achieve real-world impact. AI research and development takes place at a lightning-fast pace; there can be large shifts of capital and technological advantage overnight. Due to these power dynamics, we believe that certain governance “nodes” may be significantly more impactful or tractable than others, but we acknowledge that—again, owing to the dynamics of this field—we face many uncertainties when trying to determine which governance approaches are robustly lead to positive outcomes. This is why we have developed a methodology to steer the allocation of our resources towards approaches we have determined are the most likely to have a positive impact.

Inclusion

We believe that legitimate and sustainable governance requires bringing to the table many different perspectives. When teams lack diversity—in academic, social, ethnic, or political backgrounds—their outcomes tend to overlook critical perspectives and perpetuate socioeconomic inequalities. Ours is a field of considerable homogeneity. We make an effort to counterbalance this by engaging with underrepresented communities, working with teams and individuals representing a broad range of stakeholders—including governments, intergovernmental organizations, the private sector, academia and civil society, and by creating a professional environment in which all individuals’ identities and preferences are respected.

Innovation

We think out-of-the-box to develop solutions for operationalizing governance, putting institutional innovation at the center of our methodology. We are concerned that governments are not sufficiently equipped to mitigate the risks posed by advanced AI systems. In a field characterized by rapid capital influx and technological diffusion, many governments lack the means to measure and monitor the impact that AI systems will have on individual, sociological, and economic levels. Standing to this task will require innovation and adaptation at the institutional level—in some cases, institutions may need to adapt to new functions, or entirely new institutions may need to be formed, to support the structural changes needed for effective AI governance across sectors and geographies.

International Coordination

We believe that effective governance requires international dialogue and coordination among nations. The AI revolution is a global phenomenon. Though there are development "hot spots," the impacts of these technologies are not constrained by geographic boundaries or political borders. To effectively address the large-scale risks posed by advanced AI systems, we believe we will require an unprecedented degree of cooperation and coordination between governments and intergovernmental institutions.

Future-proofing

We orient our work to address both near- and long-term implications of AI. We aim to not only address the current needs and challenges posed by AI, but also to anticipate and prepare for its future, far-reaching impact on various aspects of society. By taking a forward-thinking approach, we strive to create solutions that will be relevant and sustainable in the long run.

Major milestones

Our network

We are proud to have worked with the following organizations: