Applications for Cohort 1 of the KAISH Introduction Program close 15/10/2026, 23:59 Erbil time.

Apply
Kurdistan AI Safety Hub Apply for Cohort 1 کوردی

Kurdistan AI Safety Hub · Erbil

Building the research capacity to make advanced AI safe, starting in Kurdistan.

We introduce students and young professionals to AI safety: how modern AI is changing the world, what can go wrong, and how they can build careers that help. Small cohorts, open discussion, and a direct line to the global AI safety community.

Cohort 112 weeksstarts at 13/10/2026Freeends at 16/10/2026

Plate IA pattern from inside a language model, woven like a kilim. Each row is a word, each cell a word it looks back at. Researchers study patterns like this to understand what AI systems are doing, one of the kinds of work we discuss in Week 5.

Why the Hub exists

Talent is everywhere. Access to this field is not.

Work on making AI systems safe is concentrated in a handful of cities. Students in Erbil, Sulaimani, Duhok and Halabja have the talent to take part, but few structured routes in: no local community, no mentors, and no one to ask where to start.

The Hub is that route, built to the standard of established international programs and grounded in our own context.

Flagship program

Applications open

KAISH Introduction Program

Twelve weeks of guided discussions: what AI safety is and why it matters, and how to grow as a professional and build a successful career.

Duration
12 weeks, 16/10/2026 to 1/1/2026
Format
Part-time and hybrid: a weekly session in Erbil, plus an online section
Weekly commitment
One discussion session
Prerequisites
None. No coding or technical background needed
Cost
Free
Cohort size
Up to 20 participants

What we do

Education and cohorts

Structured programs with a fixed cohort

Weekly facilitated discussions in small groups. Our first is the 12-week KAISH Introduction Program, and every program we design is published openly.

See the program

Technical reading groups

One paper at a time, read closely

Small groups in interpretability, evaluations, scalable oversight and governance, each led by someone who knows the paper well. The first groups open to Cohort 1 participants and invited researchers in Oct-2026.

Browse the reading library

Field building

A line to the international field

Introductions to mentors, guidance on applying to international fellowships and research programs, and talks from researchers working on these problems. We also train local facilitators so the Hub can grow beyond one cohort.

Join our team

After 12 weeks

What you’ll be able to do

Understand it

Explain in plain words how today’s AI systems are made, what they do well, and where they fail.

Think it through

Weigh the main arguments about AI risk, and say which you find convincing and why.

Present yourself

Speak and write clearly, work well in a team, and have a CV and profile ready for opportunities.

Take the next step

Leave with a personal plan for your next six months: a program to apply to, a skill to build, or a person to learn from.

Our program draws on openly published material from established AI safety programs. We are not formally affiliated with those organizations.

News and events

Cohort 1

Cohort 1 starts 16/10/2026.

Applications close 15/10/2026 at 23:59 Erbil time. The form takes about 20 minutes, and every application is read by at least two members of our team.

Stay in touch

Occasional emails about application rounds, reading groups and public talks. A few messages a term. We never share your address.