AI Safety Events & Training: 2025 week 26 update

This is a weekly newsletter listing newly announced AI safety events and training programs. Visit AISafety.com/events-and-training for the full directory of upcoming events and programs.

New events

• AI Safety x Physics Grand Challenge 25–27 July (online).
  Research hackathon from Apart, designed to be an entry point for physicists who want to tackle critical AI safety problems – allowing them to work alongside peers who share their technical background and passion for rigorous problem-solving. There will be a project track and a paper track.

• The Curve 2025 3–5 October (Berkeley, USA).
  Conference where 250 thinkers, builders, and leaders will gather to discuss where AI is going, how it will impact the world, and what to do about it. Attendees will be a curated group with a diversity of viewpoints and expertise, from frontier lab staffers to government, academia, VC, journalism, and more.

New training programs

• AI Safety Collab Course: 2025 Summer 4 August – 27 October (online).
  Comprehensive 8-week introductory program run by ENAIS, split into 2 tracks: governance (following the BlueDot curriculum) and alignment (following the AI Safety Atlas textbook). The course is aimed at people who have an interest in and are relatively new to AI safety, regardless of background.

• GovAI: Winter Fellowship 2026 19 January – 10 April 2026 (London, UK).
  3-month research fellowship with the Centre for the Governance of AI, open to early-career individuals or established professionals new to the field of AI governance. Fellows will work on a research project, deepening their knowledge of the field and forging connections with other researchers and practitioners.

Featured resource

• AI Risk: Why Care?
  An AI safety chatbot that excels at addressing hard questions and counterarguments about existential risk, with about one-third of its responses to specific questions providing genuinely great and valid arguments.