Maroc - EURASIAREVIEW.COM - A la une - 10/Sep 23:27

China’s Ambitions For Global AI Governance – Analysis

By Huw Roberts In July 2024, Shanghai hosted its annual World AI Conference, bringing together leading experts, businesses and governments to showcase cutting-edge technologies, explore future AI trends and promote global cooperation. For thefirst time, the 2024 conference foregrounded questions of global AI governance, with Shanghai’s government using the opening ceremony to issue theShanghai Declaration on Global AI Governance. The Shanghai Declaration is a short document that lacks detail beyond a high-level emphasis on maintaining AI safety and promoting inclusive global cooperation. When read in isolation, it carries little significance. But the document should be understood in the context of awider pushby the Chinese government to outline and begin implementing a vision for global AI governance. This effort was initiated in October 2023, when Chinese President Xi Jinping used his keynote speech at the opening ceremony of the Third Belt and Road Forum for International Cooperation to put forward China’s vision — the Global AI Governance Initiative.The document stresses that AI should be developed for the good of humanity, while upholding principles of mutual respect and equality. Shortly after, China proposed a resolution on enhancing international cooperation and capacity building for AI, which was unanimously adopted by the 78th session of the UN General Assembly. Like the Shanghai Declaration, the Global AI Governance Initiative and the UN resolution are vague, but common themes can be identified across these documents. Notably, China advocates for a UN-centred approach to AI governance and positions itself as a champion of the Global South by emphasising the importance of technology transfer and inclusive governance. Given China’s aim to become the world’sprimary AI innovation centreby 2030, itsleading AI research, thematurityof its domestic AI governance initiatives andits effortsto become an international ‘norm maker’ in other policy areas, it is unsurprising thatseveral commentatorsframe recent global AI governance initiatives as stemming from a position of strength. From this perspective, China is a leading AI power that is proactively trying to shape the emerging governance landscape in line with its political ideology. This argument has some truth. China is a world leading AI developer, exporter and regulator, which undoubtedly places it in an internationallyinfluential position. But a closer examination of the wider context reveals key political and technological weaknesses at the heart of China’s recent efforts to play a more assertive role in global AI governance. First, over the past five years, numerous international and intergovernmental organisations like UNESCO, the OECD and the G7 have publishedAI governance principlesand guidance documents. Most initiatives were pioneered by Western states, at times with the specific intention toexclude China. Because of this, China has played a largely marginal role in global AI governance, with someChinese commentatorsarguing that many existing initiatives are an effort by the West to pre-emptively dictate suitable values for governing AI. Second, the release of ChatGPT and other large language models in late 2022 and early 2023 caught policymakers, including those from China, off guard. This led toheightened concernsover the threat these new technologies pose topolitical stability, with China immediatelybanning ChatGPTand implementing a strictapprovals processfor domestic AI companies looking to release public-facing large language models. Third, ChatGPT’s release shifted discussions about global AI governance from a relatively fringe conversation to apolicy priorityfor global leaders, with UN Secretary General Antonio Guterres establishing the UN’sHigh Level Body on AIand British Prime Minister Rishi Sunak convening anAI Safety Summit. Combined, these factors meant that by early 2023 Chinese policymakers were increasingly concerned about the potential dangers of cutting-edge AI systems, while also facing the risk of being further sidelined in rapidly advancing global AI discussions. Accordingly, China’s recent efforts to establish and promote a vision for global AI governance should be understood as a reactionary response to technological and political developments, rather than simply being a proactive show of strength. A key issue with this reactionary vision is that rhetoric and reality do not match. While China preaches inclusivity, the actions it has taken largely focus on building partnerships to counter a perceived Western dominance in AI, be thatthrough BRICSorwith the G77. This is understandable given the geopolitical reality. Even so, China’srefusal to signa global agreement presented at the Seoul AI Safety Summit earlier this year could be interpreted as a sign of reticence to meaningfully engage in international initiatives that are not led by China or the United Nations. Such an approach would be hugely damaging for global AI governance efforts. The United Nations will certainly have a role to play in governing AI given its inclusivity and legitimacy. But the technical and general-purpose nature of AI, combined withthe ‘gridlock’ the UN faces, leaves it as an unsuitable institution for addressing many of the pressing governance issues associated with AI. For global AI governance initiatives to succeed, pragmatic engagement between China and the West in a variety of institutions is crucial. Failure to do so will see promising initiatives undermined and the fragmentation of the global AI governance ecosystem. This would be detrimental to all. About the author: Huw Roberts is Associate Fellow at RUSI and a doctoral researcher at the Oxford Internet Institute, University of Oxford. Source: This article was published by East Asia Forum