The International Artificial Intelligence Advisory Council presented the Madrid Manifesto for responsible AI this Thursday. The document defends the legitimate role of democracies "in the configuration, governance, and supervision of Artificial Intelligence, so that its benefits, opportunities, and capabilities are shared globally and aligned with human rights and democratic values".
The manifesto warns that "the self-regulation of this technology is not enough," as it "cannot ensure that it develops in line with the public interest." The text was drafted by a group of ten internationally renowned experts in areas such as technology, ethics, economics, sociology, digital governance, and cognitive sciences, in response to a mandate from the Council of Ministers in 2024.
The presentation featured the participation of the Council's president, sociologist Manuel Castells, alongside digital policy expert Francesca Bria, and the Minister for Digital Transformation and Public Function, Óscar López. Minister López thanked the Council members for their "very valuable" work, calling it "not only for Spain, but for all those around the world who see the Council as a reference model for AI advice, defense, and evaluation".
The Advisory Council alerts about new forms of colonialism in the digital age, where data appropriates personal life for exploitation. It emphasizes that AI "must be developed and governed in service of human dignity, democracy, social justice, and shared prosperity," combining competitiveness with transparency, traceability, human oversight, and accountability.
Experts assert that AI can offer "genuine opportunities for progress" in science, health, and education if implemented correctly. However, they warn of the concentration of power over AI in "few hands," which creates "fundamental challenges for democratic governance, economic equity, and technological sovereignty".
The manifesto stresses that AI is "a race to build the society most capable of leveraging AI and benefiting from it." It highlights the need to integrate fundamental rights into AI development, promoting equality, inclusion, non-discrimination, privacy, and the safety of minors. Child protection is considered "the condition that legitimizes innovation".
Digital sovereignty is identified as "a necessary condition for democracy," enabling societies to govern critical digital infrastructures according to their values. It is built through openness and cooperation, ensuring effective control over data and systems, and reflecting cultural and territorial diversity.
The Council views public capabilities as "the cornerstone of digital sovereignty," urging robust maintenance across the AI value chain to ensure democratic oversight and public interest. AI should generate "prosperity, knowledge, and well-being for present and future generations".
The manifesto calls for responsibility and transparency from those behind AI systems, demanding that liability for damages rests with "those who develop and deploy them." It requires high-risk systems to undergo "independent evaluation before deployment" and that public authorities be able to suspend systems that escape effective human control.
The document is signed by Manuel Castells, Luciano Floridi, Paul S. Adlr, Jeroen van den Hoven, Francesca Bria, Carissa Véliz, Vint Cerf, Erika Staël von Holstein, Kate Crawford, Niklas Lundblad, and Jerome A. Feldman. The Advisory Council's meeting in Madrid was inaugurated by the Secretary of State for Digitalization and AI, María González Veracruz, and closed by Minister Óscar López.
