In order for AI to bring about global benefits, cross-cultural cooperation on ethics and governance will be essential. Such cooperation will enable advances in one part of the world to be shared with other countries, and ensure that no part of society is neglected or disproportionately negatively impacted by AI. Without such cooperation, competitive pressures between countries may also lead to underinvestment in safe, ethical, and socially beneficial AI development, increasing the global risks from AI.
However, recent years have seen an increase in narratives framing the development, deployment and governance of AI as a race between AI superpowers such as the USA, Europe, and China. These narratives often go hand in hand with a perspective that there are fundamental and irresolvable differences between these regions’ priorities and values, and they threaten to seriously undermine any prospects for international cooperation.
In this new paper, we - a team of scholars from Cambridge and Beijing - make the case for the importance of cross-cultural cooperation in AI ethics and governance, and discuss how barriers to such cooperation may be overcome. We suggest that mistrust between regions is one of the biggest barriers to greater cooperation, and that such mistrust is at least partly fueled by misunderstandings and misperceptions. While real and meaningful differences between different cultures certainly exist, claims about such differences are often oversimplified, depending on unexamined concepts and entrenched assumptions. We therefore recommend that a first step towards building greater cross-cultural cooperation must be to enhance greater mutual understanding between diverse cultures and nations.
The importance of global cooperation must nonetheless be balanced against the need to allow for diverse cultural perspectives and priorities, avoiding a situation where one or two nations try to impose their values on the rest of the world. A key challenge will therefore be to identify areas where global consensus is needed, for example on safety standards, and differentiating these from areas where cross-cultural variation is important and should be protected.
We suggest that academia has a particularly important role to play in supporting cross-cultural cooperation on AI ethics and governance. Academic communities are particularly well-suited to building greater mutual understanding between regions and cultures, due to their tradition of free-flowing, international, and intercultural exchange of ideas. We make a number of practical steps that academic institutions and individual researchers can take: including translation of key papers and reports; alternating the location of key conferences; and establishing exchange programmes. We also identify several promising research topics in AI ethics and governance that would be well-suited to international collaboration, such as conducting comparative foresight exercises exploring how different cultures perceive the future impact of AI on society.
We hope this paper, also published in Chinese, can provide a starting point for many more cross-cultural collaborations and practical initiatives in AI ethics and governance. We are currently seeking to expand our work to explore a range of the research topics identified in the paper, and coauthors Yi Zeng and Sean O hEigeartaigh have launched a translation series aiming to bring influential papers and reports to a global audience.
In the words of co-author Yi Zeng: "Although the efforts on AI Ethics and Governance from various countries and organizations are established in different cultural contexts, cultural differences provide us with different perspectives and opportunities to learn from each other. As the Analects by Confucius say, Be in harmony, yet be different. Building cross-cultural mutual trust is the foundation of global harmonious development. And we may need to appreciate, learn about, and understand different perspectives during the process of shaping a global recommendation of AI governance."