LEGAL 70 GROUP

To put in common 5 guidelines to follow by all the countries that deploy AI in health systems and manage of persoal data

• Transparency, accountability, and human oversight
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Interoperability of governance approaches

AI systems are not neutral tools. Their design and deployment shape labor markets, cultural representation (biases) , access to information, and social rights. Addressing these topics is essential to ensure that AI benefits are distributed equally. Today, most AI projects are governed by procurement rules and innovation grants and there is no obligation for post-pilot continuity.

A key cross-cutting issue that remains insufficiently captured is the post-pilot continuity and lifecycle governance of AI projects.

Spain is stepping up its efforts in AI by strengthening the role of the Barcelona Supercomputing Center (BSC-CNS). It is being upgraded to function as an "AI factory," with improvements to the MareNostrum 5 supercomputer, including a new cluster dedicated specifically to AI training.

AI Dialogue can provide a neutral space for countries to align on common principles, standards, and best practices for AI governance. It helps bridge regulatory differences, reducing fragmentation and improving interoperability across jurisdictions. It can also foster trust, transparency, and knowledge-sharing between governments, industry, and other stakeholders.

• OECD AI Policy Observatory
• Global Partnership on AI (GPAI)
• UNESCO Recommendation on the Ethics of AI
• EU AI Act
• ISO
• IEEE
• G7
• G20 and NGO's sucha as AIFOD

The Governments contribute by providing regulatory frameworks and address policy bottlenecks. The Industry/Private sector contribute by offering practical insights into technical feasibility and real-world scaling. The Academia contributes by making visible the evidence-based research and long-term trend analysis. The Civil Society contributes by ensuring ethical accountability, focusing on human rights and social impact. To make the AI Dialogue more practical and easy to engage with, we should bring together governments, industry, academia, and civil society in small working groups focused on specific topics like safety or jobs, rather than broad discussions. A mix of regional and online meetings would help include more countries, especially those that are often underrepresented. Results should be shared in a public report clear, draft it with an easy-to-understand language so everyone can follow and take part.

The Global South is still clearly underrepresented, as most AI investment and development is concentrated in just a few countries, especially the United States and China. This creates an imbalance in who shapes AI systems and governance, often leaving out the needs and perspectives of other regions. To address this, it is important to actively include these countries through targeted support, such as funding participation, building regional hubs, and strengthening local AI capacity. Ensuring access to data, infrastructure, and training can help create more equal participation and allow the Global South to play a more active role in shaping AI governance.

We should use more interactive and practical formats instead of traditional panels. Small workshops focused on real case studies can help participants work together on concrete problems, while policy labs or simulations can show how different governance approaches would work in practice. Peer-learning sessions can allow countries and organizations to share real experiences, including challenges and lessons learned. Grants to cover travelling expenses should be available for underrepresented regions. Live polls or shared documents can also facilitate the discussions keep active and inclusive.

Examples of policies that promote effective AI governance are the EU AI Act and the OECD AI Principles. Both provide clear frameworks and guidelines to ensure that AI systems are developed and used in a safe, transparent, and accountable way, while still supporting innovation. UNESCO's Recommendation on the Ethics of AI is another important reference, especially for its global and inclusive perspective.