Regulate or Innovate? Governing AI Amid the Race for AI Sovereignty
The global landscape of artificial intelligence (AI) governance is undergoing a seismic shift, transitioning from collaborative oversight to competitive advancement. The intertwining of AI with national sovereignty has generated substantial resistance to effective regulation, while significant gaps in technical expertise hinder policymakers’ ability to engage meaningfully with AI challenges.
The New Sovereignty Battleground
In a short period from November 2023 to February 2025, the world witnessed a dramatic shift in AI governance. The Bletchley Declaration, signed by 28 nations, warned of potential “serious, even catastrophic harm” from advanced AI systems. However, at the Paris AI Action Summit in February 2025, French President Emmanuel Macron stated, “If we regulate before we innovate, we won’t have any innovation of our own.” This pivot towards viewing AI capabilities as essential to national power has led to the dismissal of safety concerns as mere obstacles to technological competitiveness.
The emergence of AI has reshaped perceptions of national power, particularly following the release of ChatGPT in late 2022, which ignited a global AI race. Governments have begun channeling resources towards what is often termed AI industrial policy, prioritizing acceleration of AI development over regulatory measures. The prevailing message has become clear: Innovate first, regulate later, if at all.
As nations pursue AI innovations, the interplay among states, institutions, and the private sector becomes critical in governing this technology. Failure to establish effective governance could result in significant unintended consequences. Policymakers face three notable challenges: the link between AI and sovereign ambitions, a widening expertise gap in understanding AI’s complexities, and the substantial role of private industry in AI regulation.
The Governance Deficit
Despite a surge in AI laws and regulations—over 200 at the national or supranational level—most focus on developing AI rather than effectively governing it. Prominent examples include the EU’s AI Act and China’s New Generation AI Development Plan. In the U.S., the Trump administration’s January 2025 executive order, “Removing Barriers to U.S. Leadership in AI Infrastructure,” shifted focus from safe AI development to investment and innovation. Only the EU’s AI Act adopts a comprehensive governance approach, imposing transparency and due diligence obligations on developers.
Challenge 1: Technology as National Identity
A key challenge in designing governance frameworks is the implicit—or increasingly explicit—equivalence drawn between sovereignty and technological advancement. For instance, a 2024 report from the French government linked AI directly to national sovereignty, stating, “Our lag in artificial intelligence undermines our sovereignty.” This framing elevates AI from a mere technology to a national imperative, often relegating regulation to a secondary concern.
Challenge 2: Knowledge Asymmetry
The second challenge lies in the need to demystify AI systems for effective regulation. AI experts often disagree on the definition of AI harms and the extent of AI capabilities, creating a significant expertise gap for policymakers. As political cycles and industry changes occur rapidly, regulators often lack the necessary time and understanding to navigate the complexities of AI, leading to confusion about what constitutes “AI.”
Challenge 3: Corporate Foxes in the Technological Hen House
With substantial computational resources, tech giants like Microsoft, Google, and OpenAI dominate the governance dialogue. The EU’s Digital Services Act exemplifies this dynamic, delegating compliance to industry leaders. Consequently, these companies often establish self-regulating bodies to oversee their version of responsible AI development, effectively allowing them to dictate the regulatory framework.
The Fraught Path to Global Rules
International organizations have engaged in AI governance since 2019, with frameworks such as the OECD AI Principles and the G20 AI principles. However, these principles face practical challenges, particularly conflicting national interests that hinder binding global cooperation. The U.S. and EU have significantly divergent regulatory approaches, complicating efforts for comprehensive governance.
The New Digital Divide
The uneven distribution of AI benefits creates a governance paradox. While countries like the U.S., China, and the EU possess the resources to shape AI development, nations in the Global South face substantial disadvantages, such as limited access to computing resources and insufficient AI expertise. This situation risks exacerbating existing inequalities, where the benefits of AI are concentrated among a few while the risks are disproportionately borne by the many.
Finding a Way Forward
Addressing these governance challenges requires pragmatic solutions that recognize technological realities while preserving democratic oversight. Four promising pathways emerge:
Democratic Counterweights
Effective governance necessitates counterbalancing corporate influence. By forming coalitions among universities, civil society organizations, and public interest technologists, stakeholders can provide independent technical expertise and advocate for public values in AI.
Market Incentives for Responsible AI
Public-interest AI systems can create market pressure for higher ethical standards, encouraging companies to improve their practices when ethical alternatives are available.
Risk-Based Multilateral Frameworks
Similar to historical treaties, nations can cooperate on AI governance by focusing on specific risks that threaten shared interests, gradually building trust through incremental cooperation.
Digital Solidarity Across Regions
A vision of digital solidarity could facilitate regional cooperation and equitable AI development, allowing smaller nations to participate meaningfully in the AI economy while building domestic capacity.
Beyond the False Choice
The dichotomy between innovation and regulation is false; both are essential. As nations view AI through the lens of sovereignty, the global community faces a critical governance inflection point with lasting implications for technology and power distribution. Without proactive measures, the risk of entrenching a world where a few monopolize AI benefits while imposing its risks on others becomes a stark reality.
In conclusion, the urgent need for nuanced governance frameworks that embrace both technological ambition and public protection is paramount. By crafting solutions that channel innovation responsibly, we can ensure a more equitable technological future.
