Disclosure: I used AI assistance to draft parts of this post, then reviewed and edited it before posting.
I’ve been working on a small project called AI Safety Careers: a curated job board for roles in AI safety, AI governance, policy, evaluations, red teaming, responsible AI, and related frontier AI work.
The reason I started building it is pretty simple: when I was looking through roles in this space, I found that relevant jobs were scattered across a lot of different places — company career pages, general job boards, research orgs, policy institutions, and nonprofit websites. It was also not always obvious which roles were actually relevant to AI safety or governance, and which were just generic AI/ML roles.
The current version is still early, but it includes:
- a curated job feed
- filters for category, location, remote work, and experience level
- job detail pages
- a weekly newsletter signup
- free job submissions during beta
The part I’m most unsure about is classification. Some roles are clearly in scope, like AI safety researcher, model evaluations engineer, AI governance researcher, or responsible AI policy lead. Others are much more borderline: evals infrastructure, RLHF/data platform roles, model security, trust and safety, compliance, or broader responsible AI operations.
So I’d especially appreciate feedback on a few things:
1. Are the current categories useful?
2. Which organizations should I be tracking?
3. How should borderline roles be handled?
4. What information would help candidates judge whether a role is worth applying to?
5. How can this be useful without adding more noise to the space?
The site is here:
https://aisafetycareers.com
I’m also happy to make the classification approach more transparent if that would be useful.
