Leiden Declaration
The Leiden Declaration on AI is a policy-oriented statement calling for democratic oversight of AI rather than leaving decisions solely to tech companies or technical experts. It argues that AI systems are becoming infrastructure that shapes economies, public services, information flows, and rights, so governance should be treated as a public, political issue.
Main thesis
AI governance should be accountable to democratic institutions and the public. The declaration frames AI as a socio-technical system with broad societal effects, not just a technical product.
What it advocates
In one sentence
The declaration’s core message is: AI should be governed as a public-interest technology under democratic accountability, with transparency, rights protections, independent oversight, and meaningful public participation, especially for high-impact uses.
The full text if declaration can be found here ---> DECLARATION
The Leiden Declaration on AI is a policy-oriented statement calling for democratic oversight of AI rather than leaving decisions solely to tech companies or technical experts. It argues that AI systems are becoming infrastructure that shapes economies, public services, information flows, and rights, so governance should be treated as a public, political issue.
Main thesis
AI governance should be accountable to democratic institutions and the public. The declaration frames AI as a socio-technical system with broad societal effects, not just a technical product.
What it advocates
- Democratic control and accountability: elected institutions and independent oversight bodies should have authority over significant AI deployments, especially those affecting public services, rights, labor, or access to opportunities.
- Transparency and auditability: organizations should document training data sources, model capabilities and limits, evaluation methods, and deployment contexts, and allow meaningful external scrutiny where public impact is high.
- Rights-respecting AI: systems should be designed and governed to protect privacy, due process, non-discrimination, freedom of expression, and the ability to contest automated decisions.
- Public participation: affected communities, civil-society groups, labor representatives, researchers, and journalists should have a role in shaping AI rules and evaluating deployments.
- Risk-based governance: stronger obligations should apply to high-impact or high-risk uses (e.g., policing, welfare, hiring, education, healthcare, critical infrastructure) than to low-risk applications.
- International cooperation: because AI systems and their supply chains cross borders, governments should coordinate on standards, safety practices, and enforcement while respecting democratic sovereignty.
- Concentration of power: a small number of firms controlling compute, models, data, and deployment channels can create outsized influence over public life.
- Regulation by corporate self-governance alone: voluntary commitments are viewed as insufficient when systems have significant societal impact.
- Opaque decision-making: people should not be subject to consequential AI decisions without explanations, avenues for appeal, and independent review.
In one sentence
The declaration’s core message is: AI should be governed as a public-interest technology under democratic accountability, with transparency, rights protections, independent oversight, and meaningful public participation, especially for high-impact uses.
The full text if declaration can be found here ---> DECLARATION
✦ KONSTANTINOS MICHAILIDIS ✦
━━━━━━━━━━━━━━━━━━
✧ ΚΩΝΣΤΑΝΤΙΝΟΣ ΜΙΧΑΗΛΙΔΗΣ ✧
━━━━━━━━━━━━━━━━━━
✧ ΚΩΝΣΤΑΝΤΙΝΟΣ ΜΙΧΑΗΛΙΔΗΣ ✧