The Pulse
EU Governments Adopt Dublin Declaration on Sovereign AI
The Dublin Declaration sets out seven areas for cooperation on AI safety, infrastructure, skills, sustainability and public-sector use. Ireland’s government said the declaration was agreed before the International AI Summit and adopted on O
AI.info Team · Written with AI ·
Governments put sovereign AI on a shared agenda
The Dublin Declaration on Trustworthy, Safe and Sovereign Artificial Intelligence sets out participating European Union countries’ shared intention to coordinate on AI safety, adoption, infrastructure and Europe’s ability to develop and deploy AI technologies. Ireland’s Department of Enterprise, Tourism and Employment said the declaration was agreed in the run-up to the International AI Summit in Dublin. The declaration’s text says it was adopted on October 14, 2026.
The declaration describes participating ministers as acting under Ireland’s presidency of the Council of the EU. Rather than establish a new regulatory regime, it calls for governments to build on existing EU rules, including the AI Act and the General Data Protection Regulation, while improving the practical capabilities needed to develop and use AI safely. The text says regulation alone is not enough: it also points to accessible infrastructure, interoperable data ecosystems and technical expertise.
Irish Minister of State for Trade Promotion, Artificial Intelligence and Digital Transformation Niamh Smyth described the declaration as a political commitment to cooperation. In the department’s announcement, Smyth said:
“The Dublin Declaration is a strong political signal of Member States’ shared commitment to reinforce and accelerate the important work already under way at European level.”
Safety, computing and European capacity
The declaration sets out seven areas for joint work. They include stronger safety testing and cybersecurity safeguards; wider use of trustworthy AI in public services, critical infrastructure and businesses; and support for European research, startups, scale-ups and technology providers. It also calls for organisations to retain effective control over their data, knowledge assets and AI-enabled capabilities.
Infrastructure features prominently. The participating countries call for stronger European computing capacity and data systems, along with cross-border cooperation that pools demand, computing resources and expertise. They also call for more energy-efficient AI systems and infrastructure, reflecting concern about energy use and other environmental effects as AI capacity expands.
The text links sovereignty to choice and resilience rather than isolation. It advocates open, interoperable and secure AI ecosystems, while seeking to reduce critical dependencies and strengthen supply-chain security. It also calls for AI capabilities across European languages and improved access to computing and AI opportunities in different regions, including smaller member states.
Public services and environmental costs enter the plan
Governments propose using public procurement to create demand for trustworthy, interoperable AI products and stimulate European development. The declaration also asks public administrations to build the capacity to assess, buy and deploy AI, and encourages them to share lessons and reuse suitable public-service systems across borders.
Environmental measures go beyond a general call for efficiency. The document names energy, land and water use, critical raw materials, infrastructure sharing, software development and responsible decommissioning among the areas where AI’s footprint should be managed. It also supports common standards and methods for measuring and reporting environmental impacts, and encourages transparency for consumers and users.
The declaration leaves delivery dates open
For workers and businesses, the commitments point toward more investment in education, training and workforce development, as well as broader access to computing resources. For people using public services, the participating countries’ stated aim is to encourage systems that are trustworthy, secure and usable across languages and regions. The document does not specify how much each government will spend or set deadlines for the proposed work.
The declaration’s practical impact will depend on follow-through by participating governments, EU institutions and other partners. Its published text names no delivery timetable or specific milestones for the seven areas it identifies.