The history of technological governance is at a critical crossroads. For decades, the internet governance model served as the blueprint for international cooperation in the digital sphere: a decentralized, multi-stakeholder system based on technical protocols and the free flow of information. However, the advent of Artificial Intelligence (AI) is upsetting this balance, forcing organizations like the UN to reconsider whether old tools are sufficient to tame a technology that doesn't just transport data, but generates thought and decisions.
The Internet Model: Technical Consensus and Decentralization
Internet governance, as we have known it since the 1990s, was never the work of a central government. Organizations like ICANN and the IETF shaped the web through technical standards. The core philosophy was interoperability. If systems could "talk" to each other, the network would thrive. The UN's role, through the Internet Governance Forum (IGF), was primarily advisory and facilitative, avoiding the hard legislative intervention that could fragment the web.
In contrast, AI presents a completely different architecture. While the internet is inherently decentralized, AI tends toward concentration. Power lies in the hands of a few companies with access to massive computing power and data. This means that "technical consensus" is not enough. AI governance is not just about how systems connect, but what they are allowed to "think" and how they impact human rights, labor, and national security.
The UN Intervention: The Global Digital Compact
UN Secretary-General António Guterres has made it clear that AI requires a more active and centralized approach. The Global Digital Compact (GDC), expected to be finalized at the Summit of the Future, seeks to bridge the gap. The difference here is political: while internet governance avoided ideological conflicts by focusing on infrastructure, AI governance sits at the heart of geopolitical competition.
- Sovereignty vs. Globalization: Many states view AI as a tool of national power, leading to a trend of "digital protectionism" never seen in the era of the classic internet.
- Ethical Frameworks: The need for international standards in AI Ethics requires a philosophical agreement that the internet never needed to function.
- The Role of the Global South: The UN fears that AI will widen the digital divide, turning developing nations into mere consumers of algorithms trained on Western or Chinese values.
"Internet governance was about connecting people. AI governance is about protecting humanity from its own creations," a UN official remarked.
Geopolitics and the Specter of Fragmentation
The biggest challenge for the UN is avoiding an "AI Splinternet." In the case of the internet, despite efforts by countries like China with the Great Firewall, the core TCP/IP protocol remained unified. In AI, we are already seeing the creation of two distinct ecosystems: one based on market values and individual freedoms (West) and one focusing on state control and social stability (East).
AI governance must also address the "black box" problem. On the internet, the routing of a data packet is transparent and predictable. In AI, the decisions of a deep learning model are often not fully understood even by its creators. This lack of transparency makes regulation extremely difficult, as traditional rules of liability and accountability do not easily apply.
Conclusion: Toward a Hybrid Approach
The future of AI governance will likely look like neither the ICANN model nor the International Atomic Energy Agency (IAEA) model, despite such proposals being made. A hybrid approach will be required, combining the technical flexibility of the internet with the strict legal oversight required for high-risk technologies. The UN is called upon to play the role of guarantor, ensuring that AI remains a public good rather than a weapon in the hands of a few. The success of this endeavor will determine whether the digital age leads to a new renaissance or an era of algorithmic feudalism.
