10. September 2026 ADD blog post: Habermas in the machine Democracy is best served by lively and passionate debate, which cannot be organised solely by artificial intelligence or by rational ideals of opinion formation.
19. August 2026 New ADD researcher shows why we never can seem to agree on Facebook We are speaking past each other on social media. Quite literally.
13. July 2026 Is there still room for humans in an algorithmic energy market? The electricity in your socket is likely brought to you by an algorithm.
2. June 2026 Can AI solve the problem of meetings? There can be a lot to gain if we let AI run our meetings, says new ADD researcher.
13. May 2026 Guest blog post: Unregulated AI in Politics – what Quebec Reveals About the Code of Conduct Political Parties Need Now A shared voluntary code of conduct for the use of AI by political parties offers an immediate form of protection for democratic processes.
22. April 2026 Theresia Veronika Rampisela: When fairness in algorithms is not as fair as we think What does fairness in algorithms really mean? Theresia Veronika Rampisela, PhD student ends the ADD project with a clear insight.
Impact case study reportResearch impact 14. April 2026 Release of ADD Impact Case Studies: Pathways to Research Impact As part of the ADD five-year anniversary celebration, the Knowledge Broker Unit has released a report with seven impact cases showcasing examples of how the…
7. April 2026 ADD Blogpost: Can AI work for democracy? The ADD project turns five years old. Sine Nørholm Just therefore provides insight into her research.
17. February 2026 ADD blogpost: To move beyond ‘the digital graveyard’, Denmark needs a dedicated legal framework for public-sector AI Public AI projects encounter legal barriers during implementation, which is why Denmark needs a new legal framework.
14. January 2026 ADD blogpost: The fault in AI fairness evaluation metrics – worrying flaws and steps forward In this blog post, researchers from the ADD project examine the role of AI evaluation metrics, newly found flaws, and emerging solutions.
