Students at a DAISI talk in Durham

Durham students and academics for reducing catastrophic risks from advanced AI

Whatever your expertise, you can contribute to the conversation and make a difference.

Explore the Intro FellowshipAbout

Proudly supported by

What is AI Safety?

“AI Safety” means keeping artificial intelligence from causing harm. We focus on the most catastrophic risks, up to and including human extinction, because we believe AI capabilities are advancing fast enough to make them real. Find out more.

What we do

Your starting point in AI safety

Intro Fellowship

Build your understanding of AI safety with a welcoming group of curious people. Explore the big ideas, ask questions and discover how you could help shape the future of AI.

  • Get to grips with AI alignment and governance
  • Work through the big questions in group discussions
  • Join a community that shares your curiosity
Apply for the fellowshipExplore the programme
Intro Fellowship at DAISI
No technical background needed
Viewing Sessions and Discussions at DAISI
Everyone welcome

Viewing Sessions and Discussions

Watch AI videos and documentaries together, then discuss the ideas, questions and implications for society. Everyone welcome.

Learn More
Research Project Support at DAISI
Faculty & Students

Research Project Support

We support you to pursue AI safety research. We can support applications for fellowships and funding, and advise students on dissertation and independent study projects.

Learn More
Speaker Events at DAISI
Open to all

Speaker Events

Insightful talks by leading researchers and industry professionals on AI safety and the latest research.

Learn More
Social Events at DAISI
Everyone welcome

Social Events

Relaxed socials to connect and build community.

Learn More

Book a 1-1 chat

Want to talk about AI safety, our programmes or how you could get involved? Book a 1-1 chat with Toby, our Secretary, or Sam, our Treasurer.

Book a 1-1 with TobyBook a 1-1 with Sam

Or book a chat with someone else on the team

Pizza laid out after a DAISI event

Free pizza after almost every event

Our Events

Loading calendar...

We'll update this calendar with upcoming events. You can add it to yours:

Google CalendarOutlook CalendarApple Calendar

Click on and RSVP to any event to receive email updates for that event.

See past events

Research

Sparse Autoencoders Can Learn Graded Latents for Relational Composition

Sparse Autoencoders Can Learn Graded Latents for Relational Composition

Farrell, T., Leask, P., Al Moubayed, N.

ICML 2026 Workshop
Interpretability
Ghost Marks in the Machine: A Critical Review of SynthID for Code Provenance Monitoring

Ghost Marks in the Machine: A Critical Review of SynthID for Code Provenance Monitoring

Sherratt-Cross, E., Farrell, T., Ogden, S., Ryley, O.

Apart def/acc Hackathon
Security
Probing by Analogy: Decomposing Probes into Activations for Better Interpretability and Inter-Model Generalization

Probing by Analogy: Decomposing Probes into Activations for Better Interpretability and Inter-Model Generalization

Leask, P., Al Moubayed, N.

NeurIPS 2025 Workshop
Interpretability
Order by Scale: Relative-Magnitude Relational Composition in Attention-Only Transformers

Order by Scale: Relative-Magnitude Relational Composition in Attention-Only Transformers

Farrell, T., Leask, P., Al Moubayed, N.

NeurIPS 2025 Workshop
Interpretability
View All Research

Get involved.Make a difference.

Find out how