An illustration of a street in Boston, shown six ways:
NU AI Safety is a student-led group at Northeastern. We prepare students to help make advanced AI safe, through research and through policy.
An 8-week introduction to AI safety, on a technical or a policy track.
Researchers in AI safety and interpretability, from Boston and beyond.
Support for members doing their own AI safety research.
AI safety needs more than AI researchers and software engineers. Many of its hardest problems are about policy, law, cybersecurity, biology and psychology.
We welcome undergraduate, master’s and PhD students from every program at Northeastern.
There are four ways to get involved. Anyone can join as a member today, and fellowships begin in January 2027.
Join our Slack and come to our meetings and talks. Open to everyone, with no commitment.
Join the SlackAn eight-week program on how AI systems are built and how to make them safer.
How the fellowships workAn eight-week program on how AI should be governed. No technical background required.
How the fellowships workHelp lead the group. About 3–5 hours a week, and no prior experience required.
What the board doesOpen to undergraduate, master’s and PhD students in every program.
An eight-week introduction to AI safety, taught by people who work in the field. Most Northeastern students have no AI safety course available to them, so we created one. The first fellowships begin in January 2027.
For students who want to make AI safer by working on the technology itself.
Starts from the fundamentals.
For students who want to shape how AI is governed.
No technical background required.
Both tracks lead to the same place: complete either one and you become a fellow.
Join the mailing list and we will email you before the fellowships begin.
Small groups meet for 90 minutes each week. Each session is led by a graduate student, PhD student or invited researcher, and ends with a discussion.
Complete all eight weeks to become a fellow, on either track.
Fellows are invited to research meetings and mixers with AI safety researchers at Northeastern and across Boston. We also plan to work with Northeastern labs on projects that fellows can join.
The executive board runs the group. This year we are looking for a few students to each lead one of the areas below.
Organize the technical and policy fellowships.
Plan talks and socials.
Handle communications and bring in new members.
Build connections with other AI safety groups and funders.
Any Northeastern student, at any level and in any program. We are registered as a Khoury graduate club so that graduate students can take part, but undergraduates are equally welcome, and most of our programs are designed with them in mind.
No. The policy track requires no technical background, and the technical track starts from the fundamentals. Several executive board roles are also non-technical.
The technical track focuses on making AI systems safer by working on the technology itself: how models are trained, why aligning them is hard, and interpretability. The policy track focuses on how AI is governed: the risks, how governments and companies are responding, and what good standards and laws could look like. Completing either track makes you a fellow.
One 90-minute session per week, for eight weeks.
In January 2027. Join the mailing list and we will email you before they begin.
Members join our Slack and are welcome at all open meetings and talks. There is no time commitment.
Apply using the form in the executive board section above. The role takes about 3–5 hours a week, including one board meeting, and no prior experience is required. Students on co-op are welcome to apply.
Join our Slack and tell us about your work. We would love to hear from you.
Yes, we are a Khoury College club. You can find us in Khoury’s clubs directory.
Still have a question? Ask us in the Slack or email aisafety@northeastern.edu.