The global debates on the Global Digital Compact (GDC) are now at a new stage. The third draft of GDC did not survive a silence procedure, and the Member States are now renegotiating a revised text. The two facilitators of GDC held another round of consultations on August 22 with all interested stakeholders. The impression is that all participants expressed support for the ongoing work and agreed that the United Nations should play a key role in global digital governance.

But how much of a role should the United Nations play? Should existing mechanisms be used and updated, or should new mechanisms be added? Will it follow the New York approach (where the UN Headquarter is based, representing multilateralism) or the Geneva approach (the venue of the first phase of the WSIS with active and wide participation of non-state actors, being interpreted by many as representing multistakeholder)? These debates on GDC reveal three features of global digital governance.

The first is the disputes between Member States. The differences between Member States on the most contentious issues, such as cross-border data flows, AI governance, and online content governance, are evident. Geopolitics, digital competitions, and cultural backgrounds all contribute to these disputes. This is particularly true between the digital powers who often prefer a light-handed approach to digital governance, and the less powerful nations who call for States to regulate super online platforms. Countries vary a lot in terms of the level of digital development and national conditions. It is necessary that the Internet and digital technologies be used to narrow the global digital divide rather than be used by a few digital powers to damage other countries' critical infrastructure and conduct mass surveillance.

In terms of content, what is legal in one Member State is sometimes regarded as harmful or illegal in another jurisdiction. Digitally less developed countries would like to see the appropriate framework of laws and regulations to align and oblige platforms with more duties in content moderation. There is room for GDC to lay out more principles about content governance in content genres like hate speech, war propaganda, and gender discrimination, as well as helping content platforms to shrine them in their terms of service and content moderation through establishing relevant laws and regulations and encouraging Internet industry organizations to improve industry self-discipline systems and standards.

Digital powers also compete to keep their advantages, sometimes by taking measures to obstruct or slow down other countries' technological development intentionally. This is also evident in AI governance, where there are limitations on the flow of capital and access to Large Language Models based on the country's IP addresses. There is room for GDC to set up principles suggesting what can be done by Member States or companies in digital competition, what should not be done, and what should fall into the category of Digital Commons and reduce ideology confrontation so that all countries can gain from emerging technologies. GDC can help to strengthen the value that all countries fully enjoy the right to technological development and peaceful use and share the benefits of artificial intelligence technology. We should affirm the responsibility of states to create an enabling policy and regulatory environment that protects the rights of individuals and local communities, ensuring corporate accountability and diverse and open markets.

The second is the disputes between States and non-state actors. GDC needs to bridge the disputes between State and non-state actors about how the digital space should be governed. Non-state actors are calling for the strengthening of the multistakeholder processes in digital policymaking. A multistakeholder model has been well-established in the field of Internet governance or the management of key Internet resources. This is where Internet governance is different from the regulation of telecommunications or the trade of books, whose rules are dependent on intergovernmental processes and negotiations. The lack of trust between States and non-state actors also arises from the 2013 Snowden Leaks event, where States were found untrustworthy in terms of overcollection of data beyond the need for national security.

Thirdly, these differences and disputes are being further complicated by the evolving and ever-changing nature of emerging technologies, and hence, awareness about this nature of technology is important. This is particularly the case with AI technologies, which are increasingly being used in both military and non-military areas. Understanding the nature of these technologies can constitute a challenge for the rule-makers. This is clearly a factor that makes some wording in GDC cautious due to worries about slowing down innovation. However, what is clear is that the military applications of AI, such as killer robots, constitute much more risks than its economic application. There is no need for GDC to intentionally keep its distance from the military domain by saying that it is meant for only the non-military domain in Paragraph 4.