Pour Demain: Work in Review (2025–2026)

Over the past 18 months, Pour Demain has contributed to the implementation of Europe's emerging AI governance framework through policy engagement, technical analysis and stakeholder convening. 

Alongside AI, we’ve continued to advance evidence-based biosecurity policy and support effective international development. 

This article provides an overview of our work from 2025 into 2026.

AI governance

Context: a shifting policy landscape

In recent years, AI systems have become increasingly capable and more widely deployed, creating significant opportunities while also raising new challenges for policymakers. Effective governance is essential for harnessing these benefits while managing associated risks.

Against this backdrop, Europe has moved from debating how frontier AI should be governed to implementing one of the world's first comprehensive regulatory frameworks. With the EU AI Act entering into force and the Code of Practice for general-purpose AI models finalised, attention has shifted towards enforcement, institutional capacity and ensuring that governance mechanisms can keep pace with rapid technological progress.

This evolving policy landscape has shaped much of Pour Demain's work over the past 18 months.

Shaping Europe’s AI governance framework

Pour Demain supported the development and implementation of Europe's emerging AI governance framework, with a key focus on the EU AI Act and, in particular, the Code of Practice for general-purpose AI models. 

Throughout the drafting processes, Pour Demain provided technical and policy input aimed at strengthening safety-critical elements of the framework, including proposals for mandatory external model evaluations, structured serious incident reporting, and clearer pre-deployment safety requirements. 

The final Code reflects meaningful progress in areas such as external assessments and incident reporting, while also leaving notable gaps, including limited provisions for whistleblower protections and more systematic pre-deployment transparency.

Strengthening implementation capacity

Alongside our engagement with the Code of Practice, Pour Demain published a series of policy briefings and recommendations to support the effective implementation of the AI Act and strengthen Europe's AI governance institutions.

In October 2025, we published a policymaker briefing on responding to serious AI incidents and a report arguing that the EU AI Office requires adequate resources to enforce the AI Act effectively.

In 2025, Pour Demain worked on the project Safe Artificial Intelligence for Basel-Stadt, helping critical infrastructure operators with the safe and secure deployment of AI, while feeding practical lessons into wider standardisation and legislative processes.

This focus on institutional capacity continued in April 2026, when Pour Demain joined civil society organisations, industry representatives and experts in calling on the European Commission to adequately resource the AI Office and remove administrative bottlenecks.

In 2026, Pour Demain has published a technical proposal exploring practical mechanisms for independent safety evaluations of frontier AI systems while safeguarding commercially sensitive information. We also published analysis arguing that implementation of the AI Act should account for loss-of-control risks and proposing practical measures to strengthen oversight.

Beyond the EU, Pour Demain is working with Swiss policymakers and the public to assess the potential economic impacts of advanced AI and build necessary capacity to respond to AI-enabled cyber risks. We are also working to shape the agenda of the Geneva AI Summit 2027.

Bringing together stakeholders

Convening stakeholders has been another important part of Pour Demain's work. Throughout 2025 and into 2026, we have brought together policymakers, researchers, regulators and industry representatives through multiple events and workshops. 

In March 2026, Pour Demain and SaferAI convened a roundtable at the European Parliament on Reliable, Safe & Secure AI, gathering parliamentarians, EU officials, researchers, investors and industry representatives. The meeting resulted in a joint declaration, signed by Members of the European Parliament, leading researchers including Yoshua Bengio, and industry representatives, calling for sustained European investment in reliable, safe and secure AI. 

Other highlights include a presentation at the 2025 International Association for Safe and Ethical Artificial Intelligence (IASEAI) conference, a high-level Brussels roundtable on the EU Code of Practice, and a keynote on frontier AI governance at the AI4People Summit.

Pour Demain also had several media engagements and appearances, including in Politico and Euractiv.  

Biosecurity

Alongside our contributions to AI Safety and Governance, Pour Demain continued to advance evidence-based biosecurity policy from 2021 to early 2026 (we are now shifting our resources fully to AI safety for the coming years). Much of the recent work focused on the governance of synthetic biology and nucleic acid synthesis, one of the most tractable near-term opportunities for reducing biological risk.

In March 2026, Pour Demain contributed to discussions on the forthcoming EU Biotech Act through a joint submission with 14 organisations from academia, industry and civil society. The submission sets out practical recommendations for introducing EU-wide nucleic acid synthesis screening and strengthening Europe's biosecurity framework. We also published Guidelines for Nucleic Acid Synthesis Screening in Europe. 

This built on our earlier work, including a joint biosecurity briefing published with partners such as RAND Europe and rani, as well as contributions to EU consultations on stockpiling and medical countermeasures.

Alongside these policy contributions, Pour Demain convened regular biosecurity policy diffusion meetings, bringing together leading organisations to strengthen collaboration and knowledge exchange across the field.

Global development cooperation

Pour Demain continues to support evidence-based development policy through our engagement with Coopération Globale in Switzerland and Kooperation Global e.V. in Germany.

Conclusion

Across AI, biosecurity and global development, Pour Demain's work is united by the common objective of translating evidence and scientific findings into policy advice and public debate. Whether contributing to Europe's AI governance framework, advancing biosecurity policy, or promoting more effective development cooperation, our focus remains on identifying evidence-based solutions that strengthen institutions, improve decision-making, and help policy keep pace with rapidly evolving technologies and global challenges.