Government of Catalonia

#NAME?

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

Nowadays, AI development concentrates significant power, particularly in terms of data ownership and control over algorithms. It is therefore essential to advocate for a framework that ensures strong safeguards, upholds the rule of law, and places societal benefit at its core.

In building AI capacity, it is essential to account for the specific needs of smaller companies, as well as the unique circumstances faced by firms in the Global South. As AI adoption accelerates, many companies risk losing competitiveness, potentially leading to increased market concentration in the hands of a few large players.

Governance gaps in AI—particularly in areas such as transparency, accountability, data governance, and equitable access—are already having tangible effects across Catalonia's economy and society. As AI adoption accelerates, these gaps risk undermining trust, exacerbating inequalities, and concentrating technological and economic power in a limited number of actors. From the perspective of trustworthy, ethical, and human-centric AI, one of the most significant challenges is ensuring that AI systems are aligned with fundamental rights, democratic values, and societal well-being. While regulatory frameworks are advancing, important gaps remain in their practical implementation—particularly in terms of organisational capacity, oversight mechanisms, and the ability of SMEs and public institutions to operationalise ethical and trustworthy AI in real-world settings. These challenges are directly acknowledged and addressed by the Catalonia AI 2030 Strategy, which positions the region as a European leader in responsible, ethical, and secure AI. The strategy explicitly embeds principles such as transparency, fairness, accountability, privacy, and respect for human dignity and the rule of law at the core of AI development and deployment. A key governance gap relates to the uneven distribution of capabilities. Smaller companies and less-resourced actors often lack access to data, infrastructure, and talent, limiting their ability to adopt AI responsibly. This creates risks of market concentration and dependency on large technology providers. In response, the Catalonia AI 2030 Strategy includes targeted measures to support SMEs, promote shared infrastructures, and foster public–private collaboration, thereby enabling a more inclusive AI ecosystem. At the same time, these developments open significant opportunities. Catalonia's strategic commitment to human-centric and trustworthy AI provides a strong foundation to build public trust and differentiate its ecosystem globally. Instruments such as impact assessments on fundamental rights, public registries of AI systems, and support tools for ethical compliance help translate high-level principles into actionable governance mechanisms. Moreover, by aligning innovation with ethical safeguards and societal needs, Catalonia can leverage AI as a driver of sustainable growth, improved public services, and social cohesion. The strategy's emphasis on inclusivity, territorial balance, and citizen empowerment ensures that AI adoption does not leave segments of society behind, but instead contributes to shared prosperity. In this context, addressing governance gaps is not only a matter of mitigating risks, but also a strategic lever to strengthen Catalonia's position as a leading hub for trustworthy, ethical, and human-centric AI, fully aligned with European values and regulatory frameworks.

- Promoting an overarching framework and standards that enhance international exchanges and cooperation. - Setting up a vision (a shared vision for AI). - Launching global challenges where inetrnational actors could collaborate in order to face them and propose solutions.

Building on the principles of trustworthy, ethical, and human-centric AI reflected in Catalonia's AI 2030 Strategy, the UN Global Dialogue on AI should adopt a multi-level, multi-stakeholder, and action-oriented governance model. 1. Strategic Steering (Global Level) Establish a UN High-Level Council on AI, composed of Member States, international organisations, and independent experts. This body would define strategic priorities, ensure alignment with human rights and international law, and provide political legitimacy. It should explicitly anchor AI governance in principles such as inclusivity, accountability, transparency, and societal benefit. 2. Multi-Stakeholder Advisory Mechanism Create a permanent advisory forum including private sector, academia, civil society, and technical communities. Reflecting Catalonia's collaborative approach, this body would ensure diversity of perspectives, including voices from the Global South, SMEs, and underrepresented groups, helping to reduce governance gaps and asymmetries in influence. 3. Thematic Working Groups (Operational Level) Set up thematic tracks (e.g., data governance, AI safety, public sector AI, compute access, standards and assurance). These groups would produce actionable outputs: guidelines, policy toolkits, and best practices, enabling implementation of trustworthy AI across different contexts. 4. Implementation and Capacity-Building Hub A dedicated AI Capacity and Implementation Facility should support countries—especially developing economies—in operationalising ethical AI. This includes technical assistance, training, regulatory sandboxes, and shared infrastructure, mirroring Catalonia's focus on enabling adoption beyond leading actors. 5. Monitoring, Accountability, and Transparency Introduce voluntary reporting mechanisms, peer reviews, and global observatories to track progress. Public registries of AI systems and impact assessments could be promoted as good practices to enhance accountability and trust. 6. Iterative and Adaptive Governance Ensure the Dialogue operates as a living platform, regularly updating frameworks in line with technological advances, fostering continuous learning and international coordination.

Cities, regions, smaller communities, the Global South.. Examples of how they could be included: - Establish regional AI dialogue hubs feeding into the global process - Ensure bottom-up agenda setting, not only top-down consultation - Promoting their participation (with voice if they cannot vote)

Challenges, and including student/ younger generations mobility

- Ethics-by-design and rights-based governance: The Catalonia AI 2030 Strategy embeds trustworthy, human-centric AI principles (transparency, fairness, accountability, privacy) from the outset. Example: the use of algorithmic impact assessments and ethical guidelines for public sector AI systems to ensure alignment with fundamental rights before deployment. - Public leadership through responsible AI adoption: The Catalan government acts as a lead user of AI, setting standards for transparency and accountability. Example: deployment of AI in public services (e.g., health or mobility) combined with human oversight mechanisms and emerging practices such as AI system registries to increase transparency and citizen trust. - Inclusive ecosystem and capacity building: Strong focus on enabling SMEs, research centres, and startups to participate in AI development. Example: initiatives such as the AI Hub in Barcelona, support for shared data and testing infrastructures, and targeted programmes to help smaller actors adopt AI responsibly, reducing dependency on large technology providers and fostering a more balanced ecosystem.