YRI Hackathon: AI Safety in APAC

​If you're based in APAC, or have an AI safety project idea that could help people in APAC, sign up! We're hosting a weekend of building (no experience needed!), along with guest speakers that have significant experience working on AI safety in Asia.

Form a team or compete solo for a prize pool of USD$500 - hopefully to further support your project even once the weekend comes to a close.

We're an online hackathon so work from wherever works for you, and our events will be primarily GMT +8 friendly.

For the technical folks out there, some interesting ideas might involve extending existing work (with datasets like SEA-VL (https://aclanthology.org/2025.acl-long.916/), SEATauBench (https://arxiv.org/abs/2606.28715), SEA-Guard (https://arxiv.org/abs/2602.01618), SEA-Embed (https://arxiv.org/abs/2606.03027) and many others!)

For everyone who's into governance, don't worry- a lot of what's unsolved in this region isn't technical. Some food for thought:Japan, Korea, and Singapore have AI safety institutes. Most of APAC doesn't- what's the minimum viable version for a government without the same budget? Korea's AI Basic Act came into force this January, with fines held off until 2027 - what do compliance workflows look like?

While we're an "open" hackathon with no specific categories besides a broad mandate of AIS in APAC, we'd deeply recommend you engage deeply with a problem you'd find meaningful. Our final submission format will be a PDF research report on your key findings, approach, and implications, due Sunday night. We'll judge based on the novelty of your idea, it's impacts on people in APAC, and the rigour / general quality of your research.

More info coming as we get closer to the date, and happy hacking!