We implement and maintain Decidim, a leading platform for citizen participation. Our platforms face subtle AI-generated spam, so we collaborate with public institutions and the Decidim community to develop AI-driven defenses. Our focus: using AI ethically to safeguard democracy. What models work best and are the most reliable? What infrastructure is needed? What transparency standards should apply? How do we minimize environmental impact? We're tackling these questions to ensure participation remains fair and open and we'll be glad to share insights during the conference.