Earlier this year, the AI4SE collaboration came together at the JetBrains Amsterdam office for a full day of reflection, research, and community.

The event started with a look back on our work in 2025. Participants shared highlights, lessons learned, and milestones from across the collaboration. It was a great opportunity to step back, celebrate progress, and exchange perspectives on how the field — and our research — is evolving. In the afternoon, Ali Al-kaswan and Roland Vízner gave a research talk titled From Prompt to Persist: Experiments on Autonomous CTF Agents.They presented their experiments with LLM-based autonomous agents tackling Capture The Flag cybersecurity challenges, offering insights into how these systems reason, plan, and interact with specialized security tools in complex environments.

The day continued with a planning session focused on what’s ahead for AI4SE, discussing potential research directions and collaboration opportunities for the coming year. Of course, no annual gathering would be complete without some fun — the evening wrapped up with bowling and dinner nearby, giving everyone a chance to connect informally and strengthen the community behind the collaboration.