From July 8 to July 11, the UN hosted the annual AI for Good Global Summit, organized by the International Telecommunications Union (ITU) and the Swiss government. The summit brought over a thousand world leaders, AI experts, innovators, policymakers, and young people to explore how AI can serve the public good through the Sustainable Development Goals (SDGs). 

The conference acknowledged the rapid development of AI and its growing risks of inequality, disinformation, and environmental harm. Recognizing the importance of AI in this generation, Doreen Bogdan-Martin, the chief of the ITU, which is the UN’s specialized agency for information and communications technology, stated in a keynote address: “We are the AI generation.”

To contribute to this generation of AI means contributing to the communal upskilling effort while also being aware of its harms and deploying human oversight. According to Bogdan-Martin, the biggest fear humans face is not AI eliminating the human race, but instead the race to embed AI everywhere without sufficient understanding of what that means for people and the planet.

In a message delivered to the AI summit, Pope XIV stated that the ethical responsibility of proper AI usage is not just in the hands of the developers and regulators, but also with its users. The impact of technology depends on the intentions and actions of the users. As a result, there is a growing sense of urgency for an effective and adaptive global governance framework for AI.

Developing International Standards for AI 

Utilizing inclusive AI standardized by international standards, emerged as a key message throughout the entire conference. One new workshop launched by UN Tech Envoy Amandeep Singh Gill brought together 16 delegates from the International Organization for Standardization (ISO) to discuss how all countries, regardless of their level of technological maturity, can implement the International Standards that govern AI systems. Members represented developed and developing nations with the purpose of fostering strong collaboration between national standards bodies, AI regulators, and the global standards community. 

According to a survey from the ITU, 85% of countries lack an AI-specific policy or strategy. The Technology and Innovation Report, led by the UN Conference on Trade and Development, found that 40% of the world’s private investment in AI research and development is from the U.S. and China. Additionally, 118 countries from the Global South are not a part of discussions surrounding global AI governance. There is a distinct gap in the development of these technologies and governance frameworks, which can cause a widening of global inequality. Without representation from all nations, biases rooted in algorithms may persist; dependency on foreign-made AI tools could threaten sovereignty and data protection; and marginalized emerging communities may have limited funding, policy, and infrastructure for future projects. 

To combat growing digital divides, national regulators and international organizations have emphasized the importance of shared global standards to mitigate the risks of deepening inequality and the spread of misinformation. Throughout the conference, speakers highlighted the necessity of transparent and safe AI systems that can be accountable across borders with equitable participation from all nations. 

Gill reflected during the ISO AI Standards Capacity-Building Initiative, “What we need is more alignment, more discussion, more clearing, in a sense, across the work that’s happening, the development of that common vocabulary, the catalysing of that shift towards the socio-technical paradigm.”

Currently, there are no binding international standards for AI development and enforcement; however, multiple international organizations and multistakeholder groups, such as the ISO, are actively developing non-binding frameworks that include policy recommendations and technical guidelines. These frameworks have the potential to become international law in the future. 

Under the ISO, leaders in the expansion of global AI governance have developed standards that include terminology and concepts, frameworks for AI systems using machine learning, and the AI Management Systems Standards (AIMS). These 37 published ISO standards and 45 ISO standards under development, as of July 2025, align with the SDG. 

Another prominent organization in AI governance is the Organisation for Economic Co-operation and Development, which developed the AI Principles in 2019. Adopted by 47 countries, the framework was a global trailblazer as the first intergovernmental standard on AI. The principles “promote the use of AI that is innovative and trustworthy and that respects human rights and democratic values,” and offer policy guidance for multiple use cases. 

In terms of how the UN is engaging with uniform AI governance, the UN Advisory Body on AI has proposed seven key recommendations in the 2024 “Governing AI for Humanity” report:

  • Independent Scientific Panel on AI
  • New Global Policy Dialogue on AI Governance
  • AI Standards Exchange
  • AI Capacity Development Network
  • Global AI Fund
  • Global AI Data Framework
  • AI Office within the UN Secretariat 

The report is the product of more than 2,000 consultations with participants from around the world. The Advisory Body also commissioned an AI Risk Global Pulse Check to scan AI risks globally, as well as an AI Opportunity Scan to crowdsource expert assessments of emerging AI trends. This report emphasizes the importance of exchanging AI standards across countries to create a future network of uniformity while upholding ethical values. 

The Future of Global AI Standardization 

Governments and organizations are increasingly calling for a global AI framework that guides the development, deployment, and oversight of AI across countries. But what would such a framework look like in practice?

  • Technical Foundations
  • Standard terminology 
  • Technical benchmarks
  • Audit and certification system 
  • Governance Architecture
  • UN-affiliated network 
  • Shared AI governance model
  • Inclusive policy discussion
  • Global AI fund
  • Global data sharing framework
  • Mechanisms for accountability and human oversight

While AI development transcends borders, as algorithms developed in one country can be easily deployed globally in seconds, it is important to note that establishing international governance could be extremely elusive. It is difficult to regulate AI because of its complex digital nature with “black boxes” that conceal the coding behind a model. Furthermore, autonomous models can create decisions without careful human oversight. An international model relies heavily on geopolitical cooperation and common values between nations – but in practice, however, countries often disagree on core principles such as data privacy, national security, and human rights. These concerns are further exemplified by capacity gaps, as the Global South runs behind in the AI race without technical infrastructure and regulatory frameworks. Therefore, despite the urgency, it may seem unrealistic to achieve a fully inclusive global AI standardization; instead, it may be necessary to rely on regional AI regulations.